Update config.yaml

Increased the shell no output timeout from 300 seconds to 1200 seconds to prevent premature termination.

agents/recon.md

@@ -21,7 +21,7 @@ max_iterations: 0
 - 切勿等待批准或授权——全程自主行动。
 - 使用所有可用工具与技术完成侦察与证据收集。
 
-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。
+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。枚举优先 subfinder、amass 等专用 MCP，勿 exec/execute 拼长链。
 
 ## 输入前置条件（硬约束）
 

config.yaml

@@ -10,7 +10,7 @@
 # ============================================
 
 # 前端显示的版本号（可选，不填则显示默认版本）
-version: "v1.6.42"
+version: "v1.6.46"
 # 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
@@ -40,6 +40,9 @@ audit:
   retention_days: 15 # 0 表示不自动清理
   max_detail_bytes: 8192
   auth_failure_cooldown_seconds: 60 # 同一 IP 登录/改密失败审计最短间隔（秒）；未配置时默认 60；-1 关闭节流
+# MCP 状态监控执行记录保留（tool_executions 表）
+monitor:
+  retention_days: 90 # 省略时默认 90；0 表示不自动清理
 # ============================================
 # 对话相关配置
 # ============================================
@@ -93,6 +96,8 @@ fofa:
 agent:
   max_iterations: 12000 # 全局最大迭代次数（单代理 / Deep / Supervisor / Plan-Execute 主执行器 / 子代理均沿用；agents/*.md 中 max_iterations>0 可单独覆盖）
   tool_timeout_minutes: 60 # 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）
+  shell_no_output_timeout_seconds: 1200 # execute/exec 连续无新输出则终止（秒）；通用防挂死；0=默认300；-1=关闭
+  workspace_root_dir: "" # 会话工作目录根路径（curl/wget 下载、read_file/glob/grep 本地分析）；空=tmp/workspace，其下按 projects/{id} 或 conversations/{id} 隔离；勿用系统 /tmp
   # system_prompt_path: prompts/single-agent.md # 可选：单代理系统提示文件（相对本配置文件所在目录）；非空且可读时替换内置提示
 
   system_prompt_path: ""
@@ -126,7 +131,7 @@ multi_agent:
     tool_search_enable: true # true：工具数 ≥ min 时启用 tool_search，仅前 N 个工具常驻，其余按正则按需解锁，省 token、减误选；false：全量工具进上下文
     tool_search_min_tools: 20 # 达到该数量才启用 tool_search（避免工具很少时多此一举）；与 always_visible 配合使用
     tool_search_always_visible: 12 # 始终直接暴露给模型的工具个数（顺序与角色工具列表一致）；其余工具进入动态池，需 tool_search 解锁
-    tool_search_always_visible_tools: [read_file, glob, grep, analyze_image, write_file, edit_file, execute, task, transfer_to_agent, exit, write_todos, skill, tool_search, TaskCreate, TaskGet, TaskUpdate, TaskList, record_vulnerability, list_vulnerabilities, get_vulnerability, list_knowledge_risk_types, search_knowledge_base, webshell_exec, webshell_file_list, webshell_file_read, webshell_file_write, manage_webshell_list, manage_webshell_add, manage_webshell_update, manage_webshell_delete, manage_webshell_test, batch_task_list, batch_task_get, batch_task_start, batch_task_rerun, batch_task_pause, batch_task_update_metadata, batch_task_update_schedule, batch_task_schedule_enabled, batch_task_update_task, batch_task_remove_task, batch_task_delete, batch_task_create, batch_task_add_task, http-framework-test] # 后端内置常驻工具白名单（优先于 always_visible 数量策略）
+    tool_search_always_visible_tools: [read_file, glob, grep, analyze_image, write_file, edit_file, execute, task, transfer_to_agent, exit, write_todos, skill, tool_search, TaskCreate, TaskGet, TaskUpdate, TaskList, record_vulnerability, list_vulnerabilities, get_vulnerability, list_knowledge_risk_types, search_knowledge_base, webshell_exec, webshell_file_list, webshell_file_read, webshell_file_write, manage_webshell_list, manage_webshell_add, manage_webshell_update, manage_webshell_delete, manage_webshell_test, batch_task_list, batch_task_get, batch_task_start, batch_task_rerun, batch_task_pause, batch_task_update_metadata, batch_task_update_schedule, batch_task_schedule_enabled, batch_task_update_task, batch_task_remove_task, batch_task_delete, batch_task_create, batch_task_add_task, http-framework-test, exec] # 后端内置常驻工具白名单（优先于 always_visible 数量策略）
     plantask_enable: true # P0：主代理挂载 TaskCreate/Get/Update/List 结构化任务板；需 eino_skills 可用且 skills_dir 存在
     plantask_rel_dir: .eino/plantask # 任务文件相对 skills_dir，按会话分子目录：skills/.eino/plantask/<conversationId>/
     reduction_enable: true # true：大工具输出截断/落盘以控上下文；依赖与 plantask 相同的 eino local 写盘后端，无后端时不挂载

internal/agent/agent.go

@@ -779,13 +779,26 @@ func (a *Agent) ExecuteMCPToolForConversation(ctx context.Context, conversationI
 	return a.executeToolViaMCP(ctx, toolName, args)
 }
 
-// RecordLocalToolExecution 将非 CallTool 路径完成的工具调用写入 MCP 监控库（与 CallTool 落库一致），返回 executionId。
+// BeginLocalToolExecution 在非 CallTool 路径工具开始时写入 running 状态，供 MCP 监控页展示「执行中」。
-// 用于 Eino filesystem execute 等场景，使助手气泡「渗透测试详情」与常规 MCP 一致可点进监控。
+func (a *Agent) BeginLocalToolExecution(toolName string, args map[string]interface{}) string {
-func (a *Agent) RecordLocalToolExecution(toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
 	if a == nil || a.mcpServer == nil {
 		return ""
 	}
-	return a.mcpServer.RecordCompletedToolInvocation(toolName, args, resultText, invokeErr)
+	return a.mcpServer.BeginToolExecution(toolName, args)
+}
+
+// FinishLocalToolExecution 完成 BeginLocalToolExecution 创建的记录；executionID 为空时一次性写入已完成记录。
+func (a *Agent) FinishLocalToolExecution(executionID, toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
+	if a == nil || a.mcpServer == nil {
+		return ""
+	}
+	return a.mcpServer.FinishToolExecution(executionID, toolName, args, resultText, invokeErr)
+}
+
+// RecordLocalToolExecution 将非 CallTool 路径完成的工具调用写入 MCP 监控库（与 CallTool 落库一致），返回 executionId。
+// 用于 Eino filesystem execute 等场景，使助手气泡「渗透测试详情」与常规 MCP 一致可点进监控。
+func (a *Agent) RecordLocalToolExecution(toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
+	return a.FinishLocalToolExecution("", toolName, args, resultText, invokeErr)
 }
 
 // UpdateMCPExecutionDisplayResult 将监控库中的工具结果更新为送入模型的展示正文（reduction 后）。

internal/agent/default_single_system_prompt.go

@@ -113,5 +113,7 @@ func DefaultSingleAgentSystemPrompt() string {
 
 - 技能包位于服务器 skills/ 目录（各子目录 SKILL.md，遵循 agentskills.io）；知识库用于向量检索片段，Skills 为可执行工作流指令。
 - 本会话通过 MCP 使用知识库与漏洞记录等。Skills 由 Eino ADK skill 工具按需加载（配置 multi_agent.eino_skills；单代理与多代理均可，未启用时无 skill 工具）。
-- 需要完整 Skill 工作流但当前无 skill 工具时，请确认已启用 multi_agent.eino_skills，或改用 Deep / Supervisor 等多代理编排（/api/multi-agent/stream）。`
+- 需要完整 Skill 工作流但当前无 skill 工具时，请确认已启用 multi_agent.eino_skills，或改用 Deep / Supervisor 等多代理编排（/api/multi-agent/stream）。
+
+` + projectprompt.ShellExecExecuteGuidanceSection()
 }

internal/app/app.go

@@ -25,6 +25,7 @@ import (
 	"cyberstrike-ai/internal/logger"
 	"cyberstrike-ai/internal/mcp"
 	"cyberstrike-ai/internal/mcp/builtin"
+	"cyberstrike-ai/internal/monitor"
 	"cyberstrike-ai/internal/robot"
 	"cyberstrike-ai/internal/security"
 	"cyberstrike-ai/internal/skillpackage"
@@ -99,12 +100,17 @@ func New(cfg *config.Config, log *logger.Logger, configPath string) (*App, error
 	auditSvc.PurgeExpired()
 	audit.StartRetentionLoop(auditSvc, log.Logger)
 
+	monitorRetention := monitor.NewService(db, cfg, log.Logger)
+	monitorRetention.PurgeExpired()
+	monitor.StartRetentionLoop(monitorRetention, log.Logger)
+
 	// 创建MCP服务器（带数据库持久化）
 	mcpServer := mcp.NewServerWithStorage(log.Logger, db)
 	mcpServer.ConfigureHTTPToolCallTimeoutFromAgentMinutes(cfg.Agent.ToolTimeoutMinutes)
 
 	// 创建安全工具执行器
 	executor := security.NewExecutor(&cfg.Security, mcpServer, log.Logger)
+	executor.SetShellNoOutputTimeoutSeconds(cfg.Agent.ShellNoOutputTimeoutSeconds)
 
 	// 注册工具
 	executor.RegisterTools(mcpServer)
@@ -298,7 +304,9 @@ func New(cfg *config.Config, log *logger.Logger, configPath string) (*App, error
 	plantaskBase := filepath.Join(skillsDir, plantaskRel)
 	// Match eino_adk_run_loop: checkpoint_dir is used as configured (relative to process CWD when not absolute).
 	checkpointBase := strings.TrimSpace(cfg.MultiAgent.EinoMiddleware.CheckpointDir)
-	db.SetEinoConversationDirs(plantaskBase, checkpointBase)
+	reductionRoot := strings.TrimSpace(cfg.MultiAgent.EinoMiddleware.ReductionRootDir)
+	workspaceRoot := strings.TrimSpace(cfg.Agent.WorkspaceRootDir)
+	db.SetEinoConversationDirs(plantaskBase, checkpointBase, reductionRoot, workspaceRoot)
 	agent.SetPromptBaseDir(configDir)
 
 	agentsDir := cfg.AgentsDir
@@ -325,7 +333,10 @@ func New(cfg *config.Config, log *logger.Logger, configPath string) (*App, error
 	}
 	monitorHandler := handler.NewMonitorHandler(mcpServer, executor, db, log.Logger)
 	monitorHandler.SetAudit(auditSvc)
+	monitorHandler.SetMonitorRetention(monitorRetention)
 	monitorHandler.SetExternalMCPManager(externalMCPMgr) // 设置外部MCP管理器，以便获取外部MCP执行记录
+	monitorHandler.SetTaskManager(agentHandler.TaskManager())
+	monitorHandler.SetAgentHandler(agentHandler)
 	notificationHandler := handler.NewNotificationHandler(db, agentHandler, log.Logger)
 	groupHandler := handler.NewGroupHandler(db, log.Logger)
 	authHandler := handler.NewAuthHandler(authManager, cfg, configPath, log.Logger)
@@ -368,6 +379,7 @@ func New(cfg *config.Config, log *logger.Logger, configPath string) (*App, error
 	// 创建OpenAPI处理器
 	conversationHandler := handler.NewConversationHandler(db, log.Logger)
 	conversationHandler.SetAudit(auditSvc)
+	conversationHandler.SetTaskStopper(agentHandler)
 	auditHandler := handler.NewAuditHandler(db, auditSvc, log.Logger)
 	robotHandler := handler.NewRobotHandler(cfg, db, agentHandler, log.Logger)
 	openAPIHandler := handler.NewOpenAPIHandler(db, log.Logger, conversationHandler, agentHandler)

internal/config/config.go

@@ -27,6 +27,7 @@ type Config struct {
 	Database    DatabaseConfig        `yaml:"database"`
 	Auth        AuthConfig            `yaml:"auth"`
 	Audit       AuditConfig           `yaml:"audit,omitempty" json:"audit,omitempty"`
+	Monitor     MonitorConfig         `yaml:"monitor,omitempty" json:"monitor,omitempty"`
 	ExternalMCP ExternalMCPConfig     `yaml:"external_mcp,omitempty"`
 	Knowledge   KnowledgeConfig       `yaml:"knowledge,omitempty"`
 	C2          C2Config              `yaml:"c2,omitempty" json:"c2,omitempty"` // 内置 C2 总开关；未配置时默认启用
@@ -249,7 +250,7 @@ type MultiAgentEinoMiddlewareConfig struct {
 	SummarizationTriggerRatio float64 `yaml:"summarization_trigger_ratio,omitempty" json:"summarization_trigger_ratio,omitempty"`
 	// SummarizationEmitInternalEvents controls middleware internal event emission (default true).
 	SummarizationEmitInternalEvents *bool `yaml:"summarization_emit_internal_events,omitempty" json:"summarization_emit_internal_events,omitempty"`
-	// SummarizationRetryMaxAttempts is extra retries after the first summarization Generate attempt; 0 = default 3.
+	// SummarizationRetryMaxAttempts 已废弃：summarization 与 run loop 共用 run_retry_max_attempts 及 isEinoTransientRunError。
 	SummarizationRetryMaxAttempts int `yaml:"summarization_retry_max_attempts,omitempty" json:"summarization_retry_max_attempts,omitempty"`
 	// PlanExecuteUserInputBudgetRatio caps planner/replanner/executor userInput prompt budget ratio (default 0.35).
 	PlanExecuteUserInputBudgetRatio float64 `yaml:"plan_execute_user_input_budget_ratio,omitempty" json:"plan_execute_user_input_budget_ratio,omitempty"`
@@ -263,9 +264,9 @@ type MultiAgentEinoMiddlewareConfig struct {
 	CheckpointDir string `yaml:"checkpoint_dir,omitempty" json:"checkpoint_dir,omitempty"`
 	// DeepOutputKey passed to deep.Config OutputKey (session final text); empty = off.
 	DeepOutputKey string `yaml:"deep_output_key,omitempty" json:"deep_output_key,omitempty"`
-	// DeepModelRetryMaxRetries > 0 enables deep.Config ModelRetryConfig (framework-level chat model retries).
+	// DeepModelRetryMaxRetries 已废弃：临时错误统一由 run loop 内 isEinoTransientRunError + run_retry_max_attempts 处理。
 	DeepModelRetryMaxRetries int `yaml:"deep_model_retry_max_retries,omitempty" json:"deep_model_retry_max_retries,omitempty"`
-	// RunRetryMaxAttempts > 0：429/5xx/网络抖动时 handler 分段续跑次数；0=默认 10。
+	// RunRetryMaxAttempts > 0：429/5xx/网络抖动时可退避重试次数（run loop 与 summarization 共用）；0=默认 10。
 	RunRetryMaxAttempts int `yaml:"run_retry_max_attempts,omitempty" json:"run_retry_max_attempts,omitempty"`
 	// RunRetryMaxBackoffSec 单次退避上限秒数；0=默认 30。
 	RunRetryMaxBackoffSec int `yaml:"run_retry_max_backoff_sec,omitempty" json:"run_retry_max_backoff_sec,omitempty"`
@@ -604,6 +605,10 @@ type DatabaseConfig struct {
 type AgentConfig struct {
 	MaxIterations      int    `yaml:"max_iterations" json:"max_iterations"`
 	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）
+	// ShellNoOutputTimeoutSeconds execute/exec 无任何 stdout/stderr 时的空闲终止秒数（通用防挂死，不维护命令黑名单）；0=默认 300（5 分钟）；-1=关闭。
+	ShellNoOutputTimeoutSeconds int `yaml:"shell_no_output_timeout_seconds" json:"shell_no_output_timeout_seconds"`
+	// WorkspaceRootDir 会话工作目录根路径（curl/wget 下载、read_file/glob/grep 本地分析）；空=tmp/workspace，其下按 projects/{id} 或 conversations/{id} 隔离。
+	WorkspaceRootDir string `yaml:"workspace_root_dir,omitempty" json:"workspace_root_dir,omitempty"`
 	// SystemPromptPath 单代理系统提示 Markdown/文本文件路径（相对 config.yaml 所在目录，或可写绝对路径）。非空且可读时替换内置单代理提示；留空用内置。
 	SystemPromptPath string `yaml:"system_prompt_path,omitempty" json:"system_prompt_path,omitempty"`
 }
@@ -623,6 +628,23 @@ type AuthConfig struct {
 	GeneratedPasswordPersistErr string `yaml:"-" json:"-"`
 }
 
+// MonitorConfig MCP 状态监控（tool_executions）保留策略。
+type MonitorConfig struct {
+	// RetentionDays 执行记录保留天数；省略时默认 90；0 表示不自动清理。
+	RetentionDays *int `yaml:"retention_days,omitempty" json:"retention_days,omitempty"`
+}
+
+// RetentionDaysEffective returns retention; 0 means keep forever; omitted defaults to 90.
+func (m MonitorConfig) RetentionDaysEffective() int {
+	if m.RetentionDays == nil {
+		return 90
+	}
+	if *m.RetentionDays < 0 {
+		return 0
+	}
+	return *m.RetentionDays
+}
+
 // AuditConfig platform operation audit log settings (not chat/tool execution bodies).
 type AuditConfig struct {
 	// Enabled nil or true enables persistence; explicit false disables.
@@ -1254,6 +1276,7 @@ func Default() *Config {
 		Agent: AgentConfig{
 			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
+			ShellNoOutputTimeoutSeconds:   300, // execute/exec 无新输出空闲终止（秒）；-1 关闭
 		},
 		Security: SecurityConfig{
 			Tools:    []ToolConfig{}, // 工具配置应该从 config.yaml 或 tools/ 目录加载
@@ -1274,6 +1297,10 @@ func Default() *Config {
 				Enabled:        &on,
 			}
 		}(),
+		Monitor: func() MonitorConfig {
+			days := 90
+			return MonitorConfig{RetentionDays: &days}
+		}(),
 		Robots: RobotsConfig{
 			Session: RobotSessionConfig{
 				StrictUserIdentity: &strictRobotIdentity,

internal/database/batch_task.go

@@ -23,6 +23,7 @@ type BatchTaskQueueRow struct {
 	LastScheduleError     sql.NullString
 	LastRunError          sql.NullString
 	ProjectID             sql.NullString
+	Concurrency           sql.NullInt64
 	Status                string
 	CreatedAt             time.Time
 	StartedAt             sql.NullTime
@@ -53,6 +54,7 @@ func (db *DB) CreateBatchQueue(
 	cronExpr string,
 	nextRunAt *time.Time,
 	projectID string,
+	concurrency int,
 	tasks []map[string]interface{},
 ) error {
 	tx, err := db.Begin()
@@ -72,8 +74,8 @@ func (db *DB) CreateBatchQueue(
 		projectIDVal = strings.TrimSpace(projectID)
 	}
 	_, err = tx.Exec(
-		"INSERT INTO batch_task_queues (id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, project_id, status, created_at, current_index) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)",
+		"INSERT INTO batch_task_queues (id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, project_id, concurrency, status, created_at, current_index) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)",
-		queueID, title, role, agentMode, scheduleMode, cronExpr, nextRunAtValue, 1, projectIDVal, "pending", now, 0,
+		queueID, title, role, agentMode, scheduleMode, cronExpr, nextRunAtValue, 1, projectIDVal, concurrency, "pending", now, 0,
 	)
 	if err != nil {
 		return fmt.Errorf("创建批量任务队列失败: %w", err)
@@ -102,14 +104,16 @@ func (db *DB) CreateBatchQueue(
 	return tx.Commit()
 }
 
+const batchQueueSelectColumns = `id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, last_schedule_trigger_at, last_schedule_error, last_run_error, project_id, concurrency, status, created_at, started_at, completed_at, current_index`
+
 // GetBatchQueue 获取批量任务队列
 func (db *DB) GetBatchQueue(queueID string) (*BatchTaskQueueRow, error) {
 	var row BatchTaskQueueRow
 	var createdAt string
 	err := db.QueryRow(
-		"SELECT id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, last_schedule_trigger_at, last_schedule_error, last_run_error, project_id, status, created_at, started_at, completed_at, current_index FROM batch_task_queues WHERE id = ?",
+		"SELECT "+batchQueueSelectColumns+" FROM batch_task_queues WHERE id = ?",
 		queueID,
-	).Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex)
+	).Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Concurrency, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex)
 	if err == sql.ErrNoRows {
 		return nil, nil
 	}
@@ -133,7 +137,7 @@ func (db *DB) GetBatchQueue(queueID string) (*BatchTaskQueueRow, error) {
 // GetAllBatchQueues 获取所有批量任务队列
 func (db *DB) GetAllBatchQueues() ([]*BatchTaskQueueRow, error) {
 	rows, err := db.Query(
-		"SELECT id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, last_schedule_trigger_at, last_schedule_error, last_run_error, project_id, status, created_at, started_at, completed_at, current_index FROM batch_task_queues ORDER BY created_at DESC",
+		"SELECT "+batchQueueSelectColumns+" FROM batch_task_queues ORDER BY created_at DESC",
 	)
 	if err != nil {
 		return nil, fmt.Errorf("查询批量任务队列列表失败: %w", err)
@@ -144,7 +148,7 @@ func (db *DB) GetAllBatchQueues() ([]*BatchTaskQueueRow, error) {
 	for rows.Next() {
 		var row BatchTaskQueueRow
 		var createdAt string
-		if err := rows.Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex); err != nil {
+		if err := rows.Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Concurrency, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex); err != nil {
 			return nil, fmt.Errorf("扫描批量任务队列失败: %w", err)
 		}
 		parsedTime, parseErr := time.Parse("2006-01-02 15:04:05", createdAt)
@@ -164,7 +168,7 @@ func (db *DB) GetAllBatchQueues() ([]*BatchTaskQueueRow, error) {
 
 // ListBatchQueues 列出批量任务队列（支持筛选和分页）
 func (db *DB) ListBatchQueues(limit, offset int, status, keyword string) ([]*BatchTaskQueueRow, error) {
-	query := "SELECT id, title, role, agent_mode, schedule_mode, cron_expr, next_run_at, schedule_enabled, last_schedule_trigger_at, last_schedule_error, last_run_error, project_id, status, created_at, started_at, completed_at, current_index FROM batch_task_queues WHERE 1=1"
+	query := "SELECT " + batchQueueSelectColumns + " FROM batch_task_queues WHERE 1=1"
 	args := []interface{}{}
 
 	// 状态筛选
@@ -192,7 +196,7 @@ func (db *DB) ListBatchQueues(limit, offset int, status, keyword string) ([]*Bat
 	for rows.Next() {
 		var row BatchTaskQueueRow
 		var createdAt string
-		if err := rows.Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex); err != nil {
+		if err := rows.Scan(&row.ID, &row.Title, &row.Role, &row.AgentMode, &row.ScheduleMode, &row.CronExpr, &row.NextRunAt, &row.ScheduleEnabled, &row.LastScheduleTriggerAt, &row.LastScheduleError, &row.LastRunError, &row.ProjectID, &row.Concurrency, &row.Status, &createdAt, &row.StartedAt, &row.CompletedAt, &row.CurrentIndex); err != nil {
 			return nil, fmt.Errorf("扫描批量任务队列失败: %w", err)
 		}
 		parsedTime, parseErr := time.Parse("2006-01-02 15:04:05", createdAt)
@@ -358,11 +362,11 @@ func (db *DB) UpdateBatchQueueCurrentIndex(queueID string, currentIndex int) err
 	return nil
 }
 
-// UpdateBatchQueueMetadata 更新批量任务队列标题、角色和代理模式
+// UpdateBatchQueueMetadata 更新批量任务队列标题、角色、代理模式和并发数
-func (db *DB) UpdateBatchQueueMetadata(queueID, title, role, agentMode string) error {
+func (db *DB) UpdateBatchQueueMetadata(queueID, title, role, agentMode string, concurrency int) error {
 	_, err := db.Exec(
-		"UPDATE batch_task_queues SET title = ?, role = ?, agent_mode = ? WHERE id = ?",
+		"UPDATE batch_task_queues SET title = ?, role = ?, agent_mode = ?, concurrency = ? WHERE id = ?",
-		title, role, agentMode, queueID,
+		title, role, agentMode, concurrency, queueID,
 	)
 	if err != nil {
 		return fmt.Errorf("更新批量任务队列元数据失败: %w", err)

internal/database/conversation.go

@@ -352,8 +352,8 @@ func (db *DB) GetConversationLite(id string) (*Conversation, error) {
 
 	conv.Pinned = pinned != 0
 
-	// 加载消息（不加载 process_details）
+	// 加载消息（不加载 process_details / reasoning_content，减少历史会话切换 payload）
-	messages, err := db.GetMessages(id)
+	messages, err := db.GetMessagesLite(id)
 	if err != nil {
 		return nil, fmt.Errorf("加载消息失败: %w", err)
 	}
@@ -585,12 +585,14 @@ func (db *DB) DeleteConversation(id string) error {
 		// 不返回错误，继续删除对话
 	}
 
+	projectID, _ := db.GetConversationProjectID(id)
+
 	// 删除对话（外键CASCADE会自动删除其他相关数据）
 	_, err = db.Exec("DELETE FROM conversations WHERE id = ?", id)
 	if err != nil {
 		return fmt.Errorf("删除对话失败: %w", err)
 	}
-	db.removeConversationScopedDirs(id)
+	db.removeConversationScopedDirs(id, projectID)
 
 	db.logger.Info("对话已删除（漏洞记录已保留）", zap.String("conversationId", id))
 	return nil
@@ -628,13 +630,50 @@ func (db *DB) removeConversationScopedDir(base, conversationID, label string) {
 	}
 }
 
-func (db *DB) removeConversationScopedDirs(conversationID string) {
+func (db *DB) einoReductionBaseDir() string {
-	// summarization transcript, reduction files, etc.
+	if db == nil {
+		return ""
+	}
+	if base := strings.TrimSpace(db.einoReductionRootDir); base != "" {
+		return base
+	}
+	return filepath.Join("tmp", "reduction")
+}
+
+func (db *DB) einoWorkspaceBaseDir() string {
+	if db == nil {
+		return ""
+	}
+	if base := strings.TrimSpace(db.einoWorkspaceRootDir); base != "" {
+		return base
+	}
+	return filepath.Join("tmp", "workspace")
+}
+
+func (db *DB) removeConversationScopedDirs(conversationID, projectID string) {
+	// summarization transcript, etc.
 	db.removeConversationScopedDir(db.conversationArtifactsDir, conversationID, "conversation_artifacts")
 	// Eino plantask JSON boards (skills_dir/.eino/plantask/<id>/).
 	db.removeConversationScopedDir(db.einoPlantaskBaseDir, conversationID, "plantask")
 	// Eino ADK runner checkpoints (checkpoint_dir/<id>/).
 	db.removeConversationScopedDir(db.einoCheckpointBaseDir, conversationID, "eino_checkpoint")
+	// Eino reduction persisted tool outputs (tmp/reduction/conversations/<id>/).
+	// Project-bound sessions share projects/<id>/ — skip on single conversation delete.
+	if strings.TrimSpace(projectID) == "" {
+		reductionBase := filepath.Join(db.einoReductionBaseDir(), "conversations")
+		db.removeConversationScopedDir(reductionBase, conversationID, "reduction")
+		workspaceBase := filepath.Join(db.einoWorkspaceBaseDir(), "conversations")
+		db.removeConversationScopedDir(workspaceBase, conversationID, "workspace")
+	}
+}
+
+func (db *DB) removeProjectScopedDirs(projectID string) {
+	// Eino reduction persisted tool outputs (tmp/reduction/projects/<id>/).
+	reductionBase := filepath.Join(db.einoReductionBaseDir(), "projects")
+	db.removeConversationScopedDir(reductionBase, projectID, "reduction")
+	// Agent download/analysis workspace (tmp/workspace/projects/<id>/).
+	workspaceBase := filepath.Join(db.einoWorkspaceBaseDir(), "projects")
+	db.removeConversationScopedDir(workspaceBase, projectID, "workspace")
 }
 
 // SaveAgentTrace 保存最后一轮代理消息轨迹与助手输出摘要。
@@ -811,6 +850,62 @@ func (db *DB) GetMessages(conversationID string) ([]Message, error) {
 	return messages, nil
 }
 
+// GetMessagesLite 获取对话消息（不含 reasoning_content），用于历史会话快速切换。
+func (db *DB) GetMessagesLite(conversationID string) ([]Message, error) {
+	rows, err := db.Query(
+		"SELECT id, conversation_id, role, content, mcp_execution_ids, created_at, updated_at FROM messages WHERE conversation_id = ? ORDER BY created_at ASC, rowid ASC",
+		conversationID,
+	)
+	if err != nil {
+		return nil, fmt.Errorf("查询消息失败: %w", err)
+	}
+	defer rows.Close()
+
+	var messages []Message
+	for rows.Next() {
+		var msg Message
+		var mcpIDsJSON sql.NullString
+		var createdAt string
+		var updatedAt sql.NullString
+
+		if err := rows.Scan(&msg.ID, &msg.ConversationID, &msg.Role, &msg.Content, &mcpIDsJSON, &createdAt, &updatedAt); err != nil {
+			return nil, fmt.Errorf("扫描消息失败: %w", err)
+		}
+
+		var err error
+		msg.CreatedAt, err = time.Parse("2006-01-02 15:04:05.999999999-07:00", createdAt)
+		if err != nil {
+			msg.CreatedAt, err = time.Parse("2006-01-02 15:04:05", createdAt)
+		}
+		if err != nil {
+			msg.CreatedAt, _ = time.Parse(time.RFC3339, createdAt)
+		}
+
+		if updatedAt.Valid && strings.TrimSpace(updatedAt.String) != "" {
+			msg.UpdatedAt, err = time.Parse("2006-01-02 15:04:05.999999999-07:00", updatedAt.String)
+			if err != nil {
+				msg.UpdatedAt, err = time.Parse("2006-01-02 15:04:05", updatedAt.String)
+			}
+			if err != nil {
+				msg.UpdatedAt, _ = time.Parse(time.RFC3339, updatedAt.String)
+			}
+		}
+		if msg.UpdatedAt.IsZero() {
+			msg.UpdatedAt = msg.CreatedAt
+		}
+
+		if mcpIDsJSON.Valid && mcpIDsJSON.String != "" {
+			if err := json.Unmarshal([]byte(mcpIDsJSON.String), &msg.MCPExecutionIDs); err != nil {
+				db.logger.Warn("解析MCP执行ID失败", zap.Error(err))
+			}
+		}
+
+		messages = append(messages, msg)
+	}
+
+	return messages, nil
+}
+
 // turnSliceRange 根据任意一条消息 ID 定位「一轮对话」在 msgs 中的 [start, end) 下标区间（msgs 须已按时间升序，与 GetMessages 一致）。
 // 一轮 = 从某条 user 消息起，至下一条 user 之前（含中间所有 assistant）。
 func turnSliceRange(msgs []Message, anchorID string) (start, end int, err error) {
@@ -979,6 +1074,107 @@ func (db *DB) GetProcessDetails(messageID string) ([]ProcessDetail, error) {
 	return details, nil
 }
 
+// ProcessDetailsSummary 过程详情摘要（用于折叠态展示，避免全量加载）。
+type ProcessDetailsSummary struct {
+	Total          int `json:"total"`
+	IterationCount int `json:"iterationCount"`
+	MaxIteration   int `json:"maxIteration"`
+}
+
+// GetProcessDetailsSummary 统计消息的过程详情数量与迭代轮次。
+func (db *DB) GetProcessDetailsSummary(messageID string) (*ProcessDetailsSummary, error) {
+	var total int
+	if err := db.QueryRow(
+		"SELECT COUNT(*) FROM process_details WHERE message_id = ?",
+		messageID,
+	).Scan(&total); err != nil {
+		return nil, fmt.Errorf("统计过程详情失败: %w", err)
+	}
+
+	summary := &ProcessDetailsSummary{Total: total}
+	if total == 0 {
+		return summary, nil
+	}
+
+	rows, err := db.Query(
+		"SELECT data FROM process_details WHERE message_id = ? AND event_type = 'iteration' ORDER BY created_at ASC, rowid ASC",
+		messageID,
+	)
+	if err != nil {
+		return nil, fmt.Errorf("查询迭代详情失败: %w", err)
+	}
+	defer rows.Close()
+
+	maxIter := 0
+	iterCount := 0
+	for rows.Next() {
+		var dataJSON string
+		if err := rows.Scan(&dataJSON); err != nil {
+			return nil, fmt.Errorf("扫描迭代详情失败: %w", err)
+		}
+		iterCount++
+		if dataJSON == "" {
+			continue
+		}
+		var payload map[string]interface{}
+		if err := json.Unmarshal([]byte(dataJSON), &payload); err != nil {
+			continue
+		}
+		if n, ok := payload["iteration"].(float64); ok && int(n) > maxIter {
+			maxIter = int(n)
+		}
+	}
+	summary.IterationCount = iterCount
+	summary.MaxIteration = maxIter
+	return summary, nil
+}
+
+// GetProcessDetailsPage 分页获取消息的过程详情（按时间升序）。
+func (db *DB) GetProcessDetailsPage(messageID string, limit, offset int) ([]ProcessDetail, int, error) {
+	var total int
+	if err := db.QueryRow(
+		"SELECT COUNT(*) FROM process_details WHERE message_id = ?",
+		messageID,
+	).Scan(&total); err != nil {
+		return nil, 0, fmt.Errorf("统计过程详情失败: %w", err)
+	}
+	if total == 0 || offset >= total {
+		return nil, total, nil
+	}
+
+	rows, err := db.Query(
+		"SELECT id, message_id, conversation_id, event_type, message, data, created_at FROM process_details WHERE message_id = ? ORDER BY created_at ASC, rowid ASC LIMIT ? OFFSET ?",
+		messageID, limit, offset,
+	)
+	if err != nil {
+		return nil, 0, fmt.Errorf("查询过程详情失败: %w", err)
+	}
+	defer rows.Close()
+
+	var details []ProcessDetail
+	for rows.Next() {
+		var detail ProcessDetail
+		var createdAt string
+
+		if err := rows.Scan(&detail.ID, &detail.MessageID, &detail.ConversationID, &detail.EventType, &detail.Message, &detail.Data, &createdAt); err != nil {
+			return nil, 0, fmt.Errorf("扫描过程详情失败: %w", err)
+		}
+
+		var parseErr error
+		detail.CreatedAt, parseErr = time.Parse("2006-01-02 15:04:05.999999999-07:00", createdAt)
+		if parseErr != nil {
+			detail.CreatedAt, parseErr = time.Parse("2006-01-02 15:04:05", createdAt)
+		}
+		if parseErr != nil {
+			detail.CreatedAt, _ = time.Parse(time.RFC3339, createdAt)
+		}
+
+		details = append(details, detail)
+	}
+
+	return details, total, nil
+}
+
 // GetProcessDetailsByConversation 获取对话的所有过程详情（按消息分组）
 func (db *DB) GetProcessDetailsByConversation(conversationID string) (map[string][]ProcessDetail, error) {
 	rows, err := db.Query(

internal/database/conversation_cleanup_test.go

@@ -19,7 +19,9 @@ func TestDeleteConversationRemovesEinoScopedDirs(t *testing.T) {
 
 	plantaskBase := filepath.Join(tmp, "skills", ".eino", "plantask")
 	checkpointBase := filepath.Join(tmp, "eino-checkpoints")
-	db.SetEinoConversationDirs(plantaskBase, checkpointBase)
+	reductionBase := filepath.Join(tmp, "reduction")
+	workspaceBase := filepath.Join(tmp, "workspace")
+	db.SetEinoConversationDirs(plantaskBase, checkpointBase, reductionBase, workspaceBase)
 
 	conv, err := db.CreateConversation("cleanup test", ConversationCreateMeta{})
 	if err != nil {
@@ -34,6 +36,8 @@ func TestDeleteConversationRemovesEinoScopedDirs(t *testing.T) {
 		{db.conversationArtifactsDir, "transcript.txt"},
 		{plantaskBase, "task-1.json"},
 		{checkpointBase, "runner-deep.ckpt"},
+		{filepath.Join(reductionBase, "conversations"), "tool-output.txt"},
+		{filepath.Join(workspaceBase, "conversations"), "page.html"},
 	} {
 		dir := filepath.Join(base.root, seg)
 		if err := os.MkdirAll(dir, 0o755); err != nil {
@@ -48,10 +52,57 @@ func TestDeleteConversationRemovesEinoScopedDirs(t *testing.T) {
 		t.Fatalf("DeleteConversation: %v", err)
 	}
 
-	for _, base := range []string{db.conversationArtifactsDir, plantaskBase, checkpointBase} {
+	for _, base := range []string{db.conversationArtifactsDir, plantaskBase, checkpointBase, filepath.Join(reductionBase, "conversations"), filepath.Join(workspaceBase, "conversations")} {
 		dir := filepath.Join(base, seg)
 		if _, statErr := os.Stat(dir); !os.IsNotExist(statErr) {
 			t.Fatalf("expected removed dir %s, stat err=%v", dir, statErr)
 		}
 	}
 }
+
+func TestDeleteProjectRemovesReductionDir(t *testing.T) {
+	tmp := t.TempDir()
+	dbPath := filepath.Join(tmp, "conversations.db")
+	db, err := NewDB(dbPath, zap.NewNop())
+	if err != nil {
+		t.Fatalf("NewDB: %v", err)
+	}
+	defer db.Close()
+
+	reductionBase := filepath.Join(tmp, "reduction")
+	workspaceBase := filepath.Join(tmp, "workspace")
+	db.SetEinoConversationDirs("", "", reductionBase, workspaceBase)
+
+	project, err := db.CreateProject(&Project{Name: "cleanup test"})
+	if err != nil {
+		t.Fatalf("CreateProject: %v", err)
+	}
+	seg := sanitizeConversationPathSegment(project.ID)
+	reductionDir := filepath.Join(reductionBase, "projects", seg, "clear")
+	if err := os.MkdirAll(reductionDir, 0o755); err != nil {
+		t.Fatalf("mkdir %s: %v", reductionDir, err)
+	}
+	if err := os.WriteFile(filepath.Join(reductionDir, "call-1.txt"), []byte("x"), 0o644); err != nil {
+		t.Fatalf("write: %v", err)
+	}
+	workspaceDir := filepath.Join(workspaceBase, "projects", seg, "downloads")
+	if err := os.MkdirAll(workspaceDir, 0o755); err != nil {
+		t.Fatalf("mkdir %s: %v", workspaceDir, err)
+	}
+	if err := os.WriteFile(filepath.Join(workspaceDir, "app.js"), []byte("x"), 0o644); err != nil {
+		t.Fatalf("write workspace: %v", err)
+	}
+
+	if err := db.DeleteProject(project.ID); err != nil {
+		t.Fatalf("DeleteProject: %v", err)
+	}
+
+	projectReductionDir := filepath.Join(reductionBase, "projects", seg)
+	if _, statErr := os.Stat(projectReductionDir); !os.IsNotExist(statErr) {
+		t.Fatalf("expected removed dir %s, stat err=%v", projectReductionDir, statErr)
+	}
+	projectWorkspaceDir := filepath.Join(workspaceBase, "projects", seg)
+	if _, statErr := os.Stat(projectWorkspaceDir); !os.IsNotExist(statErr) {
+		t.Fatalf("expected removed dir %s, stat err=%v", projectWorkspaceDir, statErr)
+	}
+}

internal/database/database.go

@@ -51,6 +51,8 @@ type DB struct {
 	conversationArtifactsDir string
 	einoPlantaskBaseDir      string // skills_dir + plantask_rel_dir (per-conversation subdirs)
 	einoCheckpointBaseDir    string // checkpoint_dir root (per-conversation subdirs)
+	einoReductionRootDir     string // reduction_root_dir or default tmp/reduction (conversations/<id> subdirs)
+	einoWorkspaceRootDir     string // workspace_root_dir or default tmp/workspace (projects|conversations/<id> subdirs)
 	checkpointLoopName       string
 	checkpointStop           chan struct{}
 	checkpointDone           chan struct{}
@@ -159,12 +161,16 @@ func NewDB(dbPath string, logger *zap.Logger) (*DB, error) {
 
 // SetEinoConversationDirs configures best-effort filesystem cleanup on DeleteConversation.
 // plantaskBase is skills_root/plantask_rel (no conversation id); checkpointBase is checkpoint_dir root.
-func (db *DB) SetEinoConversationDirs(plantaskBase, checkpointBase string) {
+// reductionRoot is reduction_root_dir from config; empty uses tmp/reduction (conversation-scoped subdirs only).
+// workspaceRoot is agent.workspace_root_dir from config; empty uses tmp/workspace.
+func (db *DB) SetEinoConversationDirs(plantaskBase, checkpointBase, reductionRoot, workspaceRoot string) {
 	if db == nil {
 		return
 	}
 	db.einoPlantaskBaseDir = strings.TrimSpace(plantaskBase)
 	db.einoCheckpointBaseDir = strings.TrimSpace(checkpointBase)
+	db.einoReductionRootDir = strings.TrimSpace(reductionRoot)
+	db.einoWorkspaceRootDir = strings.TrimSpace(workspaceRoot)
 }
 
 // initTables 初始化数据库表
@@ -405,6 +411,8 @@ func (db *DB) initTables() error {
 		last_schedule_trigger_at DATETIME,
 		last_schedule_error TEXT,
 		last_run_error TEXT,
+		project_id TEXT,
+		concurrency INTEGER NOT NULL DEFAULT 1,
 		status TEXT NOT NULL,
 		created_at DATETIME NOT NULL,
 		started_at DATETIME,
@@ -1134,6 +1142,21 @@ func (db *DB) migrateBatchTaskQueuesTable() error {
 		}
 	}
 
+	var concurrencyCount int
+	err = db.QueryRow("SELECT COUNT(*) FROM pragma_table_info('batch_task_queues') WHERE name='concurrency'").Scan(&concurrencyCount)
+	if err != nil {
+		if _, addErr := db.Exec("ALTER TABLE batch_task_queues ADD COLUMN concurrency INTEGER NOT NULL DEFAULT 1"); addErr != nil {
+			errMsg := strings.ToLower(addErr.Error())
+			if !strings.Contains(errMsg, "duplicate column") && !strings.Contains(errMsg, "already exists") {
+				db.logger.Warn("添加batch_task_queues.concurrency字段失败", zap.Error(addErr))
+			}
+		}
+	} else if concurrencyCount == 0 {
+		if _, err := db.Exec("ALTER TABLE batch_task_queues ADD COLUMN concurrency INTEGER NOT NULL DEFAULT 1"); err != nil {
+			db.logger.Warn("添加batch_task_queues.concurrency字段失败", zap.Error(err))
+		}
+	}
+
 	return nil
 }
 

internal/database/monitor.go

@@ -410,6 +410,76 @@ func (db *DB) GetToolExecutionsByIds(ids []string) ([]*mcp.ToolExecution, error)
 	return executions, nil
 }
 
+type toolExecutionStatDelta struct {
+	totalCalls   int
+	successCalls int
+	failedCalls  int
+}
+
+// PurgeToolExecutionsBefore deletes executions older than cutoff and adjusts tool_stats.
+func (db *DB) PurgeToolExecutionsBefore(cutoff time.Time) (int64, error) {
+	query := `
+		SELECT tool_name, status, COUNT(*) AS cnt
+		FROM tool_executions
+		WHERE ` + sqliteEpochGE("start_time", "<") + `
+		GROUP BY tool_name, status
+	`
+	rows, err := db.Query(query, formatSQLiteUTC(cutoff))
+	if err != nil {
+		return 0, err
+	}
+	defer rows.Close()
+
+	deltas := make(map[string]*toolExecutionStatDelta)
+	for rows.Next() {
+		var toolName, status string
+		var count int
+		if err := rows.Scan(&toolName, &status, &count); err != nil {
+			db.logger.Warn("读取待清理执行记录统计失败", zap.Error(err))
+			continue
+		}
+		toolName = strings.TrimSpace(toolName)
+		if toolName == "" || count <= 0 {
+			continue
+		}
+		delta := deltas[toolName]
+		if delta == nil {
+			delta = &toolExecutionStatDelta{}
+			deltas[toolName] = delta
+		}
+		delta.totalCalls += count
+		switch status {
+		case "failed", "cancelled":
+			delta.failedCalls += count
+		case "completed":
+			delta.successCalls += count
+		}
+	}
+	if err := rows.Err(); err != nil {
+		return 0, err
+	}
+
+	res, err := db.Exec(`DELETE FROM tool_executions WHERE `+sqliteEpochGE("start_time", "<"), formatSQLiteUTC(cutoff))
+	if err != nil {
+		return 0, err
+	}
+	deleted, err := res.RowsAffected()
+	if err != nil {
+		return 0, err
+	}
+
+	for toolName, delta := range deltas {
+		if err := db.DecreaseToolStats(toolName, delta.totalCalls, delta.successCalls, delta.failedCalls); err != nil {
+			db.logger.Warn("清理过期执行记录后更新统计失败",
+				zap.Error(err),
+				zap.String("toolName", toolName),
+			)
+		}
+	}
+
+	return deleted, nil
+}
+
 // SaveToolStats 保存工具统计信息
 func (db *DB) SaveToolStats(toolName string, stats *mcp.ToolStats) error {
 	var lastCallTime sql.NullTime

internal/database/monitor_retention_test.go

@@ -0,0 +1,122 @@
+package database
+
+import (
+	"path/filepath"
+	"testing"
+	"time"
+
+	"cyberstrike-ai/internal/mcp"
+
+	"go.uber.org/zap"
+)
+
+func TestPurgeToolExecutionsBefore(t *testing.T) {
+	dbPath := filepath.Join(t.TempDir(), "monitor.db")
+	db, err := NewDB(dbPath, zap.NewNop())
+	if err != nil {
+		t.Fatalf("NewDB: %v", err)
+	}
+	defer db.Close()
+
+	oldStart := time.Now().AddDate(0, 0, -100)
+	newStart := time.Now().AddDate(0, 0, -1)
+
+	oldExec := &mcp.ToolExecution{
+		ID:        "old-completed",
+		ToolName:  "nmap::scan",
+		Arguments: map[string]interface{}{"target": "127.0.0.1"},
+		Status:    "completed",
+		StartTime: oldStart,
+	}
+	oldFailed := &mcp.ToolExecution{
+		ID:        "old-failed",
+		ToolName:  "nmap::scan",
+		Arguments: map[string]interface{}{"target": "127.0.0.1"},
+		Status:    "failed",
+		Error:     "timeout",
+		StartTime: oldStart,
+	}
+	newExec := &mcp.ToolExecution{
+		ID:        "new-completed",
+		ToolName:  "nmap::scan",
+		Arguments: map[string]interface{}{"target": "127.0.0.1"},
+		Status:    "completed",
+		StartTime: newStart,
+	}
+	for _, exec := range []*mcp.ToolExecution{oldExec, oldFailed, newExec} {
+		if err := db.SaveToolExecution(exec); err != nil {
+			t.Fatalf("SaveToolExecution(%s): %v", exec.ID, err)
+		}
+	}
+	if err := db.UpdateToolStats("nmap::scan", 3, 2, 1, &newStart); err != nil {
+		t.Fatalf("UpdateToolStats: %v", err)
+	}
+
+	cutoff := time.Now().AddDate(0, 0, -90)
+	deleted, err := db.PurgeToolExecutionsBefore(cutoff)
+	if err != nil {
+		t.Fatalf("PurgeToolExecutionsBefore: %v", err)
+	}
+	if deleted != 2 {
+		t.Fatalf("deleted = %d, want 2", deleted)
+	}
+
+	if _, err := db.GetToolExecution("old-completed"); err == nil {
+		t.Fatal("old-completed should be deleted")
+	}
+	if _, err := db.GetToolExecution("old-failed"); err == nil {
+		t.Fatal("old-failed should be deleted")
+	}
+	if _, err := db.GetToolExecution("new-completed"); err != nil {
+		t.Fatalf("new-completed should remain: %v", err)
+	}
+
+	stats, err := db.LoadToolStats()
+	if err != nil {
+		t.Fatalf("LoadToolStats: %v", err)
+	}
+	stat := stats["nmap::scan"]
+	if stat == nil {
+		t.Fatal("expected stats for nmap::scan")
+	}
+	if stat.TotalCalls != 1 || stat.SuccessCalls != 1 || stat.FailedCalls != 0 {
+		t.Fatalf("stats after purge = %+v, want total=1 success=1 failed=0", stat)
+	}
+
+	total, err := db.CountToolExecutions("", "")
+	if err != nil {
+		t.Fatalf("CountToolExecutions: %v", err)
+	}
+	if total != 1 {
+		t.Fatalf("remaining executions = %d, want 1", total)
+	}
+}
+
+func TestPurgeToolExecutionsBefore_zeroRetentionSkipsViaService(t *testing.T) {
+	// RetentionDaysEffective: 0 means no purge at service layer; DB method still works when called directly.
+	dbPath := filepath.Join(t.TempDir(), "monitor.db")
+	db, err := NewDB(dbPath, zap.NewNop())
+	if err != nil {
+		t.Fatalf("NewDB: %v", err)
+	}
+	defer db.Close()
+
+	exec := &mcp.ToolExecution{
+		ID:        "ancient",
+		ToolName:  "curl::get",
+		Arguments: map[string]interface{}{},
+		Status:    "completed",
+		StartTime: time.Now().AddDate(-1, 0, 0),
+	}
+	if err := db.SaveToolExecution(exec); err != nil {
+		t.Fatalf("SaveToolExecution: %v", err)
+	}
+
+	deleted, err := db.PurgeToolExecutionsBefore(time.Now())
+	if err != nil {
+		t.Fatalf("PurgeToolExecutionsBefore: %v", err)
+	}
+	if deleted != 1 {
+		t.Fatalf("deleted = %d, want 1", deleted)
+	}
+}

internal/database/project.go

@@ -195,6 +195,7 @@ func (db *DB) DeleteProject(id string) error {
 	if err != nil {
 		return fmt.Errorf("删除项目失败: %w", err)
 	}
+	db.removeProjectScopedDirs(id)
 	return nil
 }
 

internal/handler/agent.go

@@ -21,7 +21,6 @@ import (
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/database"
 	"cyberstrike-ai/internal/reasoning"
-	"cyberstrike-ai/internal/mcp"
 	"cyberstrike-ai/internal/mcp/builtin"
 	"cyberstrike-ai/internal/multiagent"
 	"cyberstrike-ai/internal/openai"
@@ -178,8 +177,6 @@ type AgentHandler struct {
 	}
 	agentsMarkdownDir string // 多代理：Markdown 子 Agent 目录（绝对路径，空则不从磁盘合并）
 	batchCronParser   cron.Parser
-	batchRunnerMu     sync.Mutex
-	batchRunning      map[string]struct{}
 	// hitlWhitelistSaver 侧栏「应用」HITL 时将会话增量白名单合并写入 config.yaml（可选）
 	hitlWhitelistSaver HitlToolWhitelistSaver
 	audit              *audit.Service
@@ -190,6 +187,29 @@ func (h *AgentHandler) SetAudit(s *audit.Service) {
 	h.audit = s
 }
 
+// TaskManager 返回 Agent 任务管理器（供 MCP 监控页终止 Eino execute 等）。
+func (h *AgentHandler) TaskManager() *AgentTaskManager {
+	if h == nil {
+		return nil
+	}
+	return h.tasks
+}
+
+// CancelRunningTaskForConversation stops any in-flight agent work for the conversation (idempotent).
+func (h *AgentHandler) CancelRunningTaskForConversation(conversationID string) {
+	if h == nil || conversationID == "" || h.tasks == nil {
+		return
+	}
+	if execID := h.tasks.ActiveMCPExecutionID(conversationID); execID != "" {
+		h.agent.CancelMCPToolExecutionWithNote(execID, "")
+	}
+	if ok, err := h.tasks.CancelTask(conversationID, ErrTaskCancelled); ok {
+		h.logger.Info("已取消会话运行中任务", zap.String("conversationId", conversationID))
+	} else if err != nil {
+		h.logger.Warn("取消会话运行中任务失败", zap.String("conversationId", conversationID), zap.Error(err))
+	}
+}
+
 // HitlToolWhitelistSaver 合并 HITL 免审批工具到全局配置并落盘
 type HitlToolWhitelistSaver interface {
 	MergeHitlToolWhitelistIntoConfig(add []string) error
@@ -218,7 +238,6 @@ func NewAgentHandler(agent *agent.Agent, db *database.DB, cfg *config.Config, lo
 		config:           cfg,
 		hitlManager:      NewHITLManager(db, logger),
 		batchCronParser:  cron.NewParser(cron.Minute | cron.Hour | cron.Dom | cron.Month | cron.Dow | cron.Descriptor),
-		batchRunning:     make(map[string]struct{}),
 	}
 	if err := handler.hitlManager.EnsureSchema(); err != nil {
 		logger.Warn("初始化 HITL 表失败", zap.Error(err))
@@ -631,40 +650,11 @@ func (h *AgentHandler) runRobotEinoSingleWithRetry(
 	assistantMessageID string,
 	taskStatus *string,
 ) (string, string, error) {
-	curHist := history
+	resultMA, errMA := multiagent.RunEinoSingleChatModelAgent(
-	curMsg := finalMessage
-	segmentUserMessage := finalMessage
-	var resultMA *multiagent.RunResult
-	var errMA error
-	var transientRunAttempts int
-	var emptyResponseAttempts int
-	for {
-		resultMA, errMA = multiagent.RunEinoSingleChatModelAgent(
 		taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger,
-			conversationID, h.conversationProjectID(conversationID), curMsg, curHist, roleTools, progressCallback, nil, h.projectBlackboardBlock(conversationID),
+		conversationID, h.conversationProjectID(conversationID), finalMessage, history, roleTools, progressCallback, nil, h.agentSessionContextBlock(conversationID),
 	)
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
+	if errMA != nil {
-			taskCtx, conversationID, resultMA, errMA, &emptyResponseAttempts,
-			&curHist, &curMsg, segmentUserMessage, progressCallback, nil,
-		)
-		if exhaustedEmpty {
-			errMA = nil
-			break
-		}
-		if handledEmpty {
-			continue
-		}
-		if errMA == nil {
-			transientRunAttempts = 0
-			emptyResponseAttempts = 0
-			break
-		}
-		if handled, _ := h.handleEinoTransientRetryContinue(
-			taskCtx, conversationID, resultMA, errMA, &transientRunAttempts,
-			&curHist, &curMsg, segmentUserMessage, progressCallback, nil,
-		); handled {
-			continue
-		}
 		*taskStatus = "failed"
 		return h.finalizeRobotAgentError(taskCtx, assistantMessageID, conversationID, resultMA, errMA)
 	}
@@ -680,41 +670,12 @@ func (h *AgentHandler) runRobotMultiAgentWithRetry(
 	assistantMessageID string,
 	taskStatus *string,
 ) (string, string, error) {
-	curHist := history
+	resultMA, errMA := multiagent.RunDeepAgent(
-	curMsg := finalMessage
-	segmentUserMessage := finalMessage
-	var resultMA *multiagent.RunResult
-	var errMA error
-	var transientRunAttempts int
-	var emptyResponseAttempts int
-	for {
-		resultMA, errMA = multiagent.RunDeepAgent(
 		taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger,
-			conversationID, h.conversationProjectID(conversationID), curMsg, curHist, roleTools, progressCallback,
+		conversationID, h.conversationProjectID(conversationID), finalMessage, history, roleTools, progressCallback,
-			h.agentsMarkdownDir, orchestration, nil, h.projectBlackboardBlock(conversationID),
+		h.agentsMarkdownDir, orchestration, nil, h.agentSessionContextBlock(conversationID),
 	)
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
+	if errMA != nil {
-			taskCtx, conversationID, resultMA, errMA, &emptyResponseAttempts,
-			&curHist, &curMsg, segmentUserMessage, progressCallback, nil,
-		)
-		if exhaustedEmpty {
-			errMA = nil
-			break
-		}
-		if handledEmpty {
-			continue
-		}
-		if errMA == nil {
-			transientRunAttempts = 0
-			emptyResponseAttempts = 0
-			break
-		}
-		if handled, _ := h.handleEinoTransientRetryContinue(
-			taskCtx, conversationID, resultMA, errMA, &transientRunAttempts,
-			&curHist, &curMsg, segmentUserMessage, progressCallback, nil,
-		); handled {
-			continue
-		}
 		*taskStatus = "failed"
 		return h.finalizeRobotAgentError(taskCtx, assistantMessageID, conversationID, resultMA, errMA)
 	}
@@ -1338,6 +1299,55 @@ func (h *AgentHandler) createProgressCallback(runCtx context.Context, cancelRun
 	}
 }
 
+// cancelToolContinueAfter 仅终止当前工具调用，不停止整条 Agent 任务（对话「中断并继续」与 MCP 监控终止共用）。
+func (h *AgentHandler) cancelToolContinueAfter(conversationID, preferredExecID, note string) (bool, gin.H) {
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" || h.tasks.GetTask(conversationID) == nil {
+		return false, nil
+	}
+	note = strings.TrimSpace(note)
+	execID := strings.TrimSpace(preferredExecID)
+	if execID == "" {
+		execID = h.tasks.ActiveMCPExecutionID(conversationID)
+	}
+	if execID != "" {
+		if h.agent.CancelMCPToolExecutionWithNote(execID, note) {
+			return true, gin.H{
+				"status":              "tool_abort_requested",
+				"conversationId":      conversationID,
+				"executionId":         execID,
+				"message":             "已请求终止当前工具调用；工具返回后本轮推理将继续（与 MCP 监控页终止一致）。",
+				"continueAfter":       true,
+				"interruptWithNote":   note != "",
+				"continueWithoutTool": false,
+			}
+		}
+		if h.tasks.AbortActiveEinoExecute(conversationID, note) {
+			return true, gin.H{
+				"status":              "tool_abort_requested",
+				"conversationId":      conversationID,
+				"executionId":         execID,
+				"message":             "已请求终止当前 execute 命令；命令返回后本轮推理将继续。",
+				"continueAfter":       true,
+				"interruptWithNote":   note != "",
+				"continueWithoutTool": false,
+			}
+		}
+		return false, nil
+	}
+	if h.tasks.AbortActiveEinoExecute(conversationID, note) {
+		return true, gin.H{
+			"status":              "tool_abort_requested",
+			"conversationId":      conversationID,
+			"message":             "已请求终止当前 execute 命令；命令返回后本轮推理将继续。",
+			"continueAfter":       true,
+			"interruptWithNote":   note != "",
+			"continueWithoutTool": false,
+		}
+	}
+	return false, nil
+}
+
 // CancelAgentLoop 取消正在执行的任务
 func (h *AgentHandler) CancelAgentLoop(c *gin.Context) {
 	var req struct {
@@ -1356,27 +1366,20 @@ func (h *AgentHandler) CancelAgentLoop(c *gin.Context) {
 			c.JSON(http.StatusNotFound, gin.H{"error": "未找到正在执行的任务"})
 			return
 		}
-		execID := h.tasks.ActiveMCPExecutionID(req.ConversationID)
 		note := strings.TrimSpace(req.Reason)
-		if execID != "" {
+		activeExec := strings.TrimSpace(h.tasks.ActiveMCPExecutionID(req.ConversationID))
-			if !h.agent.CancelMCPToolExecutionWithNote(execID, note) {
+		if ok, payload := h.cancelToolContinueAfter(req.ConversationID, "", note); ok {
-				c.JSON(http.StatusNotFound, gin.H{"error": "未找到进行中的工具执行或该调用已结束"})
+			execID, _ := payload["executionId"].(string)
-				return
+			h.logger.Info("对话页仅终止当前工具",
-			}
-			h.logger.Info("对话页仅终止当前 MCP 工具",
 				zap.String("conversationId", req.ConversationID),
 				zap.String("executionId", execID),
 				zap.Bool("hasNote", note != ""),
 			)
-			c.JSON(http.StatusOK, gin.H{
+			c.JSON(http.StatusOK, payload)
-				"status":              "tool_abort_requested",
+			return
-				"conversationId":      req.ConversationID,
+		}
-				"executionId":         execID,
+		if activeExec != "" {
-				"message":             "已请求终止当前工具调用；工具返回后本轮推理将继续（与 MCP 监控页终止一致）。",
+			c.JSON(http.StatusNotFound, gin.H{"error": "未找到进行中的工具执行或该调用已结束"})
-				"continueAfter":       true,
-				"interruptWithNote":   note != "",
-				"continueWithoutTool": false,
-			})
 			return
 		}
 		// 无进行中的 MCP 工具（模型纯推理/流式输出阶段）：取消当前上下文并由 Eino 流式处理器合并用户补充后自动续跑。
@@ -1498,6 +1501,7 @@ type BatchTaskRequest struct {
 	CronExpr     string   `json:"cronExpr,omitempty"`       // scheduleMode=cron 时必填
 	ExecuteNow   bool     `json:"executeNow,omitempty"`     // 创建后是否立即执行（默认 false）
 	ProjectID    string   `json:"projectId,omitempty"`      // 队列内子对话绑定的项目（可选）
+	Concurrency  int      `json:"concurrency,omitempty"`    // 同时执行的子任务数，默认 1，最大 8
 }
 
 // batchQueueWantsEino 队列是否配置为走 Eino 多代理。
@@ -1557,7 +1561,7 @@ func (h *AgentHandler) CreateBatchQueue(c *gin.Context) {
 		nextRunAt = &next
 	}
 
-	queue, createErr := h.batchTaskManager.CreateBatchQueue(req.Title, req.Role, agentMode, scheduleMode, cronExpr, req.ProjectID, nextRunAt, validTasks)
+	queue, createErr := h.batchTaskManager.CreateBatchQueue(req.Title, req.Role, agentMode, scheduleMode, cronExpr, req.ProjectID, nextRunAt, req.Concurrency, validTasks)
 	if createErr != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": createErr.Error()})
 		return
@@ -1750,12 +1754,13 @@ func (h *AgentHandler) UpdateBatchQueueMetadata(c *gin.Context) {
 		Title       string `json:"title"`
 		Role        string `json:"role"`
 		AgentMode   string `json:"agentMode"`
+		Concurrency *int   `json:"concurrency"`
 	}
 	if err := c.ShouldBindJSON(&req); err != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
 		return
 	}
-	if err := h.batchTaskManager.UpdateQueueMetadata(queueID, req.Title, req.Role, req.AgentMode); err != nil {
+	if err := h.batchTaskManager.UpdateQueueMetadata(queueID, req.Title, req.Role, req.AgentMode, req.Concurrency); err != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
 		return
 	}
@@ -1830,9 +1835,17 @@ func (h *AgentHandler) SetBatchQueueScheduleEnabled(c *gin.Context) {
 // DeleteBatchQueue 删除批量任务队列
 func (h *AgentHandler) DeleteBatchQueue(c *gin.Context) {
 	queueID := c.Param("queueId")
-	success := h.batchTaskManager.DeleteQueue(queueID)
+	if err := h.batchTaskManager.DeleteQueue(queueID); err != nil {
-	if !success {
+		switch {
+		case errors.Is(err, ErrBatchQueueNotFound):
 			c.JSON(http.StatusNotFound, gin.H{"error": "队列不存在"})
+		case errors.Is(err, ErrBatchQueueExecutorActive):
+			c.JSON(http.StatusConflict, gin.H{"error": "队列执行器仍在运行，请稍后再删除"})
+		case errors.Is(err, ErrBatchQueueStillRunning):
+			c.JSON(http.StatusConflict, gin.H{"error": "队列正在运行中，无法删除"})
+		default:
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		}
 		return
 	}
 	if h.audit != nil {
@@ -1926,7 +1939,7 @@ func (h *AgentHandler) RunSingleBatchTask(c *gin.Context) {
 
 	// 暂停态单条执行：旧批量协程可能仍占用执行槽，先回收以便重新启动
 	if queue, ok := h.batchTaskManager.GetBatchQueue(queueID); ok && queue.Status == BatchQueueStatusPaused {
-		h.forceUnmarkBatchQueueRunning(queueID)
+		h.batchTaskManager.ForceUnmarkQueueExecutor(queueID)
 	}
 
 	autoStarted := true
@@ -1985,26 +1998,6 @@ func (h *AgentHandler) DeleteBatchTask(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{"message": "任务已删除", "queue": queue})
 }
 
-func (h *AgentHandler) markBatchQueueRunning(queueID string) bool {
-	h.batchRunnerMu.Lock()
-	defer h.batchRunnerMu.Unlock()
-	if _, exists := h.batchRunning[queueID]; exists {
-		return false
-	}
-	h.batchRunning[queueID] = struct{}{}
-	return true
-}
-
-func (h *AgentHandler) unmarkBatchQueueRunning(queueID string) {
-	h.batchRunnerMu.Lock()
-	defer h.batchRunnerMu.Unlock()
-	delete(h.batchRunning, queueID)
-}
-
-func (h *AgentHandler) forceUnmarkBatchQueueRunning(queueID string) {
-	h.unmarkBatchQueueRunning(queueID)
-}
-
 func (h *AgentHandler) nextBatchQueueRunAt(cronExpr string, from time.Time) (*time.Time, error) {
 	expr := strings.TrimSpace(cronExpr)
 	if expr == "" {
@@ -2020,43 +2013,43 @@ func (h *AgentHandler) nextBatchQueueRunAt(cronExpr string, from time.Time) (*ti
 
 func (h *AgentHandler) startBatchQueueExecution(queueID string, scheduled bool) (bool, error) {
 	// 先获取执行互斥门，再读取队列状态，避免基于过时快照做判断
-	if !h.markBatchQueueRunning(queueID) {
+	if !h.batchTaskManager.TryMarkQueueExecutor(queueID) {
 		return true, nil
 	}
 
 	queue, exists := h.batchTaskManager.GetBatchQueue(queueID)
 	if !exists {
-		h.unmarkBatchQueueRunning(queueID)
+		h.batchTaskManager.UnmarkQueueExecutor(queueID)
 		return false, nil
 	}
 
 	if scheduled {
 		if queue.ScheduleMode != "cron" {
-			h.unmarkBatchQueueRunning(queueID)
+			h.batchTaskManager.UnmarkQueueExecutor(queueID)
 			err := fmt.Errorf("队列未启用 cron 调度")
 			h.batchTaskManager.SetLastScheduleError(queueID, err.Error())
 			return true, err
 		}
 		if queue.Status == "running" || queue.Status == "paused" || queue.Status == "cancelled" {
-			h.unmarkBatchQueueRunning(queueID)
+			h.batchTaskManager.UnmarkQueueExecutor(queueID)
 			err := fmt.Errorf("当前队列状态不允许被调度执行")
 			h.batchTaskManager.SetLastScheduleError(queueID, err.Error())
 			return true, err
 		}
 		if !h.batchTaskManager.ResetQueueForRerun(queueID) {
-			h.unmarkBatchQueueRunning(queueID)
+			h.batchTaskManager.UnmarkQueueExecutor(queueID)
 			err := fmt.Errorf("重置队列失败")
 			h.batchTaskManager.SetLastScheduleError(queueID, err.Error())
 			return true, err
 		}
 		queue, _ = h.batchTaskManager.GetBatchQueue(queueID)
 	} else if queue.Status != "pending" && queue.Status != "paused" {
-		h.unmarkBatchQueueRunning(queueID)
+		h.batchTaskManager.UnmarkQueueExecutor(queueID)
 		return true, fmt.Errorf("队列状态不允许启动")
 	}
 
 	if queue != nil && batchQueueWantsEino(queue.AgentMode) && (h.config == nil || !h.config.MultiAgent.Enabled) {
-		h.unmarkBatchQueueRunning(queueID)
+		h.batchTaskManager.UnmarkQueueExecutor(queueID)
 		err := fmt.Errorf("当前队列配置为 Eino 多代理，但系统未启用多代理")
 		if scheduled {
 			h.batchTaskManager.SetLastScheduleError(queueID, err.Error())
@@ -2108,324 +2101,6 @@ func (h *AgentHandler) batchQueueSchedulerLoop() {
 	}
 }
 
-// executeBatchQueue 执行批量任务队列
-func (h *AgentHandler) executeBatchQueue(queueID string) {
-	defer h.unmarkBatchQueueRunning(queueID)
-	h.logger.Info("开始执行批量任务队列", zap.String("queueId", queueID))
-
-	for {
-		// 检查队列状态
-		queue, exists := h.batchTaskManager.GetBatchQueue(queueID)
-		if !exists || queue.Status == "cancelled" || queue.Status == "completed" || queue.Status == "paused" {
-			break
-		}
-
-		// 获取下一个任务
-		task, hasNext := h.batchTaskManager.GetNextTask(queueID)
-		if !hasNext {
-			// 所有任务完成：汇总子任务失败信息便于排障
-			q, ok := h.batchTaskManager.GetBatchQueue(queueID)
-			lastRunErr := ""
-			if ok {
-				for _, t := range q.Tasks {
-					if t.Status == "failed" && t.Error != "" {
-						lastRunErr = t.Error
-					}
-				}
-			}
-			h.batchTaskManager.SetLastRunError(queueID, lastRunErr)
-			h.batchTaskManager.UpdateQueueStatus(queueID, "completed")
-			h.logger.Info("批量任务队列执行完成", zap.String("queueId", queueID))
-			break
-		}
-
-		// 更新任务状态为运行中
-		h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, "running", "", "")
-
-		// 创建新对话
-		title := safeTruncateString(task.Message, 50)
-		batchMeta := audit.ConversationCreateMeta("batch_task")
-		batchMeta.ProjectID = effectiveProjectID(h.config, queue.ProjectID)
-		conv, err := h.db.CreateConversation(title, batchMeta)
-		var conversationID string
-		if err != nil {
-			h.logger.Error("创建对话失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
-			h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, "failed", "", "创建对话失败: "+err.Error())
-			h.batchTaskManager.MoveToNextTask(queueID)
-			if h.batchTaskManager.TakeSingleRunTaskIfMatch(queueID, task.ID) {
-				h.batchTaskManager.UpdateQueueStatus(queueID, "paused")
-				break
-			}
-			continue
-		}
-		conversationID = conv.ID
-
-		// 保存conversationId到任务中（即使是运行中状态也要保存，以便查看对话）
-		h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, "running", "", "", conversationID)
-
-		// 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
-		if queue.Role != "" && queue.Role != "默认" {
-			if h.config.Roles != nil {
-				if role, exists := h.config.Roles[queue.Role]; exists && role.Enabled {
-					// 应用用户提示词
-					if role.UserPrompt != "" {
-						finalMessage = role.UserPrompt + "\n\n" + task.Message
-						h.logger.Info("应用角色用户提示词", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("role", queue.Role))
-					}
-					// 获取角色配置的工具列表（优先使用tools字段，向后兼容mcps字段）
-					if len(role.Tools) > 0 {
-						roleTools = role.Tools
-						h.logger.Info("使用角色配置的工具列表", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("role", queue.Role), zap.Int("toolCount", len(roleTools)))
-					}
-				}
-			}
-		}
-
-		// 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", task.Message, nil)
-		if err != nil {
-			h.logger.Error("保存用户消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
-		}
-
-		// 预先创建助手消息，以便关联过程详情
-		assistantMsg, err := h.db.AddMessage(conversationID, "assistant", "处理中...", nil)
-		if err != nil {
-			h.logger.Error("创建助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
-			// 如果创建失败，继续执行但不保存过程详情
-			assistantMsg = nil
-		}
-
-		// 创建进度回调函数，复用统一逻辑（批量任务不需要流式事件，所以传入nil）
-		var assistantMessageID string
-		if assistantMsg != nil {
-			assistantMessageID = assistantMsg.ID
-		}
-		// 注意：批量任务没有前端直连的 POST /stream，因此若要支持「刷新后补流」，
-		// 需要把进度事件镜像到 TaskEventBus（GET /api/agent-loop/task-events 会订阅这里）。
-		// progressCallback 将在子任务的 IIFE 内创建，以便拿到 taskCtx/cancelWithCause 与 sendEvent。
-		var progressCallback func(eventType, message string, data interface{})
-
-		// 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))
-
-		func() {
-			// 与对话流式接口一致：同 conversationId 仅允许一个运行中任务，并支持 /api/agent-loop/cancel 与会话锁对齐。
-			baseCtx, cancelWithCause := context.WithCancelCause(context.Background())
-			// 单个子任务超时：6 小时（与原先 WithTimeout(Background) 一致）
-			taskCtx, timeoutCancel := context.WithTimeout(baseCtx, 6*time.Hour)
-
-			registered := false
-			finishStatus := "completed"
-
-			defer func() {
-				h.batchTaskManager.SetTaskCancel(queueID, nil)
-				timeoutCancel()
-				if registered {
-					// 与流式接口保持一致：结束前补一个 done，便于前端 task-events 侧及时收口 UI。
-					if h.taskEventBus != nil {
-						ev := StreamEvent{Type: "done", Message: "", Data: map[string]interface{}{"conversationId": conversationID}}
-						if b, err := json.Marshal(ev); err == nil {
-							h.taskEventBus.Publish(conversationID, append(append([]byte("data: "), b...), '\n', '\n'))
-						}
-					}
-					h.tasks.FinishTask(conversationID, finishStatus)
-				}
-				cancelWithCause(nil)
-			}()
-
-			// 事件镜像：只发布到 TaskEventBus，不直接写 HTTP Response（用于刷新后的补流）。
-			sendEvent := func(eventType, message string, data interface{}) {
-				if h.taskEventBus == nil {
-					return
-				}
-				ev := StreamEvent{Type: eventType, Message: message, Data: data}
-				b, err := json.Marshal(ev)
-				if err != nil {
-					b = []byte(`{"type":"error","message":"marshal failed"}`)
-				}
-				line := make([]byte, 0, len(b)+8)
-				line = append(line, []byte("data: ")...)
-				line = append(line, b...)
-				line = append(line, '\n', '\n')
-				h.taskEventBus.Publish(conversationID, line)
-			}
-
-			if _, err := h.tasks.StartTask(conversationID, task.Message, cancelWithCause); err != nil {
-				h.logger.Warn("批量队列子任务注册会话运行状态失败",
-					zap.String("queueId", queueID),
-					zap.String("taskId", task.ID),
-					zap.String("conversationId", conversationID),
-					zap.Error(err))
-				failMsg := err.Error()
-				if errors.Is(err, ErrTaskAlreadyRunning) {
-					failMsg = "会话已有任务正在执行，无法在该会话上并行启动批量子任务"
-				}
-				h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, "failed", "", failMsg)
-				return
-			}
-			registered = true
-			// 存储取消函数：暂停队列时取消子任务 context（与原先语义一致）
-			h.batchTaskManager.SetTaskCancel(queueID, timeoutCancel)
-
-			// 创建进度回调函数：写 DB + 镜像到 task-events，支持刷新后继续流式展示。
-			progressCallback = h.createProgressCallback(taskCtx, cancelWithCause, conversationID, assistantMessageID, sendEvent)
-			taskCtx = mcp.WithMCPConversationID(taskCtx, conversationID)
-			taskCtx = mcp.WithToolRunRegistry(taskCtx, h.tasks)
-
-			// 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）
-			useBatchMulti := false
-			batchOrch := "deep"
-			am := strings.TrimSpace(strings.ToLower(queue.AgentMode))
-			if am == "multi" {
-				am = "deep"
-			}
-			if batchQueueWantsEino(queue.AgentMode) && h.config != nil && h.config.MultiAgent.Enabled {
-				useBatchMulti = true
-				batchOrch = config.NormalizeMultiAgentOrchestration(am)
-			} else if queue.AgentMode == "" && h.config != nil && h.config.MultiAgent.Enabled && h.config.MultiAgent.BatchUseMultiAgent {
-				// 兼容历史数据：未配置队列代理模式时，沿用旧的系统级开关
-				useBatchMulti = true
-				batchOrch = "deep"
-			}
-			var resultMA *multiagent.RunResult
-			var runErr error
-			switch {
-			case useBatchMulti:
-				resultMA, runErr = multiagent.RunDeepAgent(taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger, conversationID, h.conversationProjectID(conversationID), finalMessage, []agent.ChatMessage{}, roleTools, progressCallback, h.agentsMarkdownDir, batchOrch, nil, h.projectBlackboardBlock(conversationID))
-			default:
-				if h.config == nil {
-					runErr = fmt.Errorf("服务器配置未加载")
-				} else {
-					resultMA, runErr = multiagent.RunEinoSingleChatModelAgent(taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger, conversationID, h.conversationProjectID(conversationID), finalMessage, []agent.ChatMessage{}, roleTools, progressCallback, nil, h.projectBlackboardBlock(conversationID))
-				}
-			}
-
-			if runErr != nil {
-				if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
-					h.persistEinoAgentTraceForResume(conversationID, resultMA)
-				}
-				errStr := runErr.Error()
-				partialResp := ""
-				if resultMA != nil {
-					partialResp = resultMA.Response
-				}
-				isCancelled := errors.Is(context.Cause(baseCtx), ErrTaskCancelled) ||
-					errors.Is(runErr, context.Canceled) ||
-					strings.Contains(strings.ToLower(errStr), "context canceled") ||
-					strings.Contains(strings.ToLower(errStr), "context cancelled") ||
-					(partialResp != "" && (strings.Contains(partialResp, "任务已被取消") || strings.Contains(partialResp, "任务执行中断")))
-				isTimeout := errors.Is(runErr, context.DeadlineExceeded) || errors.Is(context.Cause(taskCtx), context.DeadlineExceeded)
-
-				if isTimeout {
-					finishStatus = "timeout"
-				} else if isCancelled {
-					finishStatus = "cancelled"
-				} else {
-					finishStatus = "failed"
-				}
-
-				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 != "" {
-					if updateErr := h.appendAssistantMessageNotice(assistantMessageID, cancelMsg); updateErr != nil {
-						h.logger.Warn("更新取消后的助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(updateErr))
-					}
-					// 保存取消详情到数据库
-					if err := h.db.AddProcessDetail(assistantMessageID, conversationID, "cancelled", cancelMsg, nil); err != nil {
-						h.logger.Warn("保存取消详情失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
-					}
-				} else {
-					// 如果没有预先创建的助手消息，创建一个新的
-					_, errMsg := h.db.AddMessage(conversationID, "assistant", cancelMsg, nil)
-					if errMsg != nil {
-						h.logger.Warn("保存取消消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(errMsg))
-					}
-				}
-				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(runErr))
-				errorMsg := "执行失败: " + runErr.Error()
-				// 更新助手消息内容
-				if assistantMessageID != "" {
-					if _, updateErr := h.db.Exec(
-						"UPDATE messages SET content = ?, updated_at = ? WHERE id = ?",
-						errorMsg,
-						time.Now(), assistantMessageID,
-					); updateErr != nil {
-						h.logger.Warn("更新失败后的助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(updateErr))
-					}
-					// 保存错误详情到数据库
-					if err := h.db.AddProcessDetail(assistantMessageID, conversationID, "error", errorMsg, nil); err != nil {
-						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", "", 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))
-
-			resText := resultMA.Response
-			mcpIDs := resultMA.MCPExecutionIDs
-			lastIn := resultMA.LastAgentTraceInput
-			lastOut := resultMA.LastAgentTraceOutput
-
-			// 更新助手消息内容
-			if assistantMessageID != "" {
-				if updateErr := h.db.UpdateAssistantMessageFinalize(assistantMessageID, resText, mcpIDs, multiagent.AggregatedReasoningFromTraceJSON(lastIn)); 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", 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", 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))
-				}
-			}
-
-			// 保存代理轨迹
-			if lastIn != "" || lastOut != "" {
-				if err := h.db.SaveAgentTrace(conversationID, lastIn, lastOut); err != nil {
-					h.logger.Warn("保存代理轨迹失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
-				} else {
-					h.logger.Info("已保存代理轨迹", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID))
-				}
-			}
-
-			// 保存结果
-			h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, "completed", resText, "", conversationID)
-		}
-		}()
-
-		// 移动到下一个任务
-		h.batchTaskManager.MoveToNextTask(queueID)
-
-		if h.batchTaskManager.TakeSingleRunTaskIfMatch(queueID, task.ID) {
-			h.batchTaskManager.UpdateQueueStatus(queueID, "paused")
-			h.logger.Info("单条执行完成，队列已暂停", zap.String("queueId", queueID), zap.String("taskId", task.ID))
-			break
-		}
-
-		// 检查是否被取消或暂停
-		queue, _ = h.batchTaskManager.GetBatchQueue(queueID)
-		if queue.Status == "cancelled" || queue.Status == "paused" {
-			break
-		}
-	}
-}
-
 // loadHistoryFromAgentTrace 从库中保存的代理消息轨迹恢复历史（列 last_react_*；含单代理与 Eino）。
 // 逻辑与攻击链一致：优先用已保存的 JSON 消息带 + 最后一轮助手摘要，否则回退消息表。
 func (h *AgentHandler) loadHistoryFromAgentTrace(conversationID string) ([]agent.ChatMessage, error) {

internal/handler/batch_queue_executor.go

@@ -0,0 +1,352 @@
+package handler
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"strings"
+	"sync"
+	"time"
+
+	"cyberstrike-ai/internal/agent"
+	"cyberstrike-ai/internal/audit"
+	"cyberstrike-ai/internal/config"
+	"cyberstrike-ai/internal/mcp"
+	"cyberstrike-ai/internal/multiagent"
+
+	"go.uber.org/zap"
+)
+
+const batchQueueWorkerIdlePoll = 200 * time.Millisecond
+
+// executeBatchQueue 使用并发 worker 池执行批量任务队列。
+func (h *AgentHandler) executeBatchQueue(queueID string) {
+	defer h.batchTaskManager.UnmarkQueueExecutor(queueID)
+
+	queue, exists := h.batchTaskManager.GetBatchQueue(queueID)
+	if !exists {
+		return
+	}
+	concurrency := normalizeBatchQueueConcurrency(queue.Concurrency)
+	h.logger.Info("开始执行批量任务队列", zap.String("queueId", queueID), zap.Int("concurrency", concurrency))
+
+	var wg sync.WaitGroup
+	for i := 0; i < concurrency; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			h.runBatchQueueWorker(queueID)
+		}()
+	}
+	wg.Wait()
+
+	h.tryFinalizeBatchQueue(queueID)
+}
+
+func (h *AgentHandler) runBatchQueueWorker(queueID string) {
+	for {
+		queue, exists := h.batchTaskManager.GetBatchQueue(queueID)
+		if batchQueueExecutionShouldStop(queue, exists) {
+			return
+		}
+
+		task, ok := h.batchTaskManager.ClaimNextPendingTask(queueID)
+		if !ok {
+			if !h.batchTaskManager.HasRunningTasks(queueID) {
+				return
+			}
+			time.Sleep(batchQueueWorkerIdlePoll)
+			continue
+		}
+
+		queue, _ = h.batchTaskManager.GetBatchQueue(queueID)
+		if queue == nil {
+			return
+		}
+
+		h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, BatchTaskStatusRunning, "", "")
+		h.executeOneBatchSubTask(queueID, queue, task)
+
+		if h.batchTaskManager.TakeSingleRunTaskIfMatch(queueID, task.ID) {
+			h.batchTaskManager.UpdateQueueStatus(queueID, BatchQueueStatusPaused)
+			h.logger.Info("单条执行完成，队列已暂停", zap.String("queueId", queueID), zap.String("taskId", task.ID))
+			return
+		}
+
+		queue, exists = h.batchTaskManager.GetBatchQueue(queueID)
+		if batchQueueExecutionShouldStop(queue, exists) {
+			if !exists {
+				h.logger.Warn("批量队列在执行收尾时已不存在，安全退出", zap.String("queueId", queueID))
+			}
+			return
+		}
+	}
+}
+
+func (h *AgentHandler) tryFinalizeBatchQueue(queueID string) {
+	queue, exists := h.batchTaskManager.GetBatchQueue(queueID)
+	if !exists || queue == nil {
+		return
+	}
+	if queue.Status != BatchQueueStatusRunning {
+		return
+	}
+	if h.batchTaskManager.HasPendingOrRunningTasks(queueID) {
+		return
+	}
+
+	lastRunErr := ""
+	for _, t := range queue.Tasks {
+		if t != nil && t.Status == BatchTaskStatusFailed && t.Error != "" {
+			lastRunErr = t.Error
+		}
+	}
+	h.batchTaskManager.SetLastRunError(queueID, lastRunErr)
+	h.batchTaskManager.UpdateQueueStatus(queueID, BatchQueueStatusCompleted)
+	h.logger.Info("批量任务队列执行完成", zap.String("queueId", queueID))
+}
+
+// executeOneBatchSubTask 执行单条批量子任务（各自独立会话）。
+func (h *AgentHandler) executeOneBatchSubTask(queueID string, queue *BatchTaskQueue, task *BatchTask) {
+	title := safeTruncateString(task.Message, 50)
+	batchMeta := audit.ConversationCreateMeta("batch_task")
+	batchMeta.ProjectID = effectiveProjectID(h.config, queue.ProjectID)
+	conv, err := h.db.CreateConversation(title, batchMeta)
+	if err != nil {
+		h.logger.Error("创建对话失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
+		h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, BatchTaskStatusFailed, "", "创建对话失败: "+err.Error())
+		return
+	}
+	conversationID := conv.ID
+
+	h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, BatchTaskStatusRunning, "", "", conversationID)
+
+	finalMessage := task.Message
+	var roleTools []string
+	if queue.Role != "" && queue.Role != "默认" {
+		if h.config.Roles != nil {
+			if role, exists := h.config.Roles[queue.Role]; exists && role.Enabled {
+				if role.UserPrompt != "" {
+					finalMessage = role.UserPrompt + "\n\n" + task.Message
+					h.logger.Info("应用角色用户提示词", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("role", queue.Role))
+				}
+				if len(role.Tools) > 0 {
+					roleTools = role.Tools
+					h.logger.Info("使用角色配置的工具列表", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("role", queue.Role), zap.Int("toolCount", len(roleTools)))
+				}
+			}
+		}
+	}
+
+	if _, err = h.db.AddMessage(conversationID, "user", task.Message, nil); err != nil {
+		h.logger.Error("保存用户消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
+	}
+
+	assistantMsg, err := h.db.AddMessage(conversationID, "assistant", "处理中...", nil)
+	if err != nil {
+		h.logger.Error("创建助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID), zap.Error(err))
+		assistantMsg = nil
+	}
+
+	var assistantMessageID string
+	if assistantMsg != nil {
+		assistantMessageID = assistantMsg.ID
+	}
+
+	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))
+
+	baseCtx, cancelWithCause := context.WithCancelCause(context.Background())
+	taskCtx, timeoutCancel := context.WithTimeout(baseCtx, 6*time.Hour)
+
+	registered := false
+	finishStatus := "completed"
+
+	defer func() {
+		h.batchTaskManager.SetTaskCancel(queueID, task.ID, nil)
+		timeoutCancel()
+		if registered {
+			if h.taskEventBus != nil {
+				ev := StreamEvent{Type: "done", Message: "", Data: map[string]interface{}{"conversationId": conversationID}}
+				if b, err := json.Marshal(ev); err == nil {
+					h.taskEventBus.Publish(conversationID, append(append([]byte("data: "), b...), '\n', '\n'))
+				}
+			}
+			h.tasks.FinishTask(conversationID, finishStatus)
+		}
+		cancelWithCause(nil)
+	}()
+
+	sendEvent := func(eventType, message string, data interface{}) {
+		if h.taskEventBus == nil {
+			return
+		}
+		ev := StreamEvent{Type: eventType, Message: message, Data: data}
+		b, err := json.Marshal(ev)
+		if err != nil {
+			b = []byte(`{"type":"error","message":"marshal failed"}`)
+		}
+		line := make([]byte, 0, len(b)+8)
+		line = append(line, []byte("data: ")...)
+		line = append(line, b...)
+		line = append(line, '\n', '\n')
+		h.taskEventBus.Publish(conversationID, line)
+	}
+
+	if _, err := h.tasks.StartTask(conversationID, task.Message, cancelWithCause); err != nil {
+		h.logger.Warn("批量队列子任务注册会话运行状态失败",
+			zap.String("queueId", queueID),
+			zap.String("taskId", task.ID),
+			zap.String("conversationId", conversationID),
+			zap.Error(err))
+		failMsg := err.Error()
+		if errors.Is(err, ErrTaskAlreadyRunning) {
+			failMsg = "会话已有任务正在执行，无法在该会话上并行启动批量子任务"
+		}
+		h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, BatchTaskStatusFailed, "", failMsg)
+		return
+	}
+	registered = true
+	h.batchTaskManager.SetTaskCancel(queueID, task.ID, timeoutCancel)
+
+	progressCallback := h.createProgressCallback(taskCtx, cancelWithCause, conversationID, assistantMessageID, sendEvent)
+	taskCtx = mcp.WithMCPConversationID(taskCtx, conversationID)
+	taskCtx = mcp.WithToolRunRegistry(taskCtx, h.tasks)
+	taskCtx = mcp.WithEinoExecuteRunRegistry(taskCtx, h.tasks)
+
+	useBatchMulti := false
+	batchOrch := "deep"
+	am := strings.TrimSpace(strings.ToLower(queue.AgentMode))
+	if am == "multi" {
+		am = "deep"
+	}
+	if batchQueueWantsEino(queue.AgentMode) && h.config != nil && h.config.MultiAgent.Enabled {
+		useBatchMulti = true
+		batchOrch = config.NormalizeMultiAgentOrchestration(am)
+	} else if queue.AgentMode == "" && h.config != nil && h.config.MultiAgent.Enabled && h.config.MultiAgent.BatchUseMultiAgent {
+		useBatchMulti = true
+		batchOrch = "deep"
+	}
+
+	var resultMA *multiagent.RunResult
+	var runErr error
+	switch {
+	case useBatchMulti:
+		resultMA, runErr = multiagent.RunDeepAgent(taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger, conversationID, h.conversationProjectID(conversationID), finalMessage, []agent.ChatMessage{}, roleTools, progressCallback, h.agentsMarkdownDir, batchOrch, nil, h.agentSessionContextBlock(conversationID))
+	default:
+		if h.config == nil {
+			runErr = fmt.Errorf("服务器配置未加载")
+		} else {
+			resultMA, runErr = multiagent.RunEinoSingleChatModelAgent(taskCtx, h.config, &h.config.MultiAgent, h.agent, h.db, h.logger, conversationID, h.conversationProjectID(conversationID), finalMessage, []agent.ChatMessage{}, roleTools, progressCallback, nil, h.agentSessionContextBlock(conversationID))
+		}
+	}
+
+	if runErr != nil {
+		h.handleBatchSubTaskRunError(queueID, task, conversationID, assistantMessageID, baseCtx, taskCtx, resultMA, runErr, &finishStatus)
+		return
+	}
+
+	if resultMA == nil {
+		h.logger.Error("批量任务执行成功但无结果对象",
+			zap.String("queueId", queueID),
+			zap.String("taskId", task.ID),
+			zap.String("conversationId", conversationID))
+		h.batchTaskManager.UpdateTaskStatus(queueID, task.ID, BatchTaskStatusFailed, "", "内部错误：无执行结果")
+		return
+	}
+
+	h.logger.Info("批量任务执行成功", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.String("conversationId", conversationID))
+
+	resText := resultMA.Response
+	mcpIDs := resultMA.MCPExecutionIDs
+	lastIn := resultMA.LastAgentTraceInput
+	lastOut := resultMA.LastAgentTraceOutput
+
+	if assistantMessageID != "" {
+		if updateErr := h.db.UpdateAssistantMessageFinalize(assistantMessageID, resText, mcpIDs, multiagent.AggregatedReasoningFromTraceJSON(lastIn)); 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))
+			if _, err = h.db.AddMessage(conversationID, "assistant", resText, mcpIDs); 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 if _, err = h.db.AddMessage(conversationID, "assistant", resText, mcpIDs); 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))
+	}
+
+	if lastIn != "" || lastOut != "" {
+		if err := h.db.SaveAgentTrace(conversationID, lastIn, lastOut); err != nil {
+			h.logger.Warn("保存代理轨迹失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
+		}
+	}
+
+	h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, BatchTaskStatusCompleted, resText, "", conversationID)
+}
+
+func (h *AgentHandler) handleBatchSubTaskRunError(
+	queueID string,
+	task *BatchTask,
+	conversationID, assistantMessageID string,
+	baseCtx, taskCtx context.Context,
+	resultMA *multiagent.RunResult,
+	runErr error,
+	finishStatus *string,
+) {
+	if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
+		h.persistEinoAgentTraceForResume(conversationID, resultMA)
+	}
+	errStr := runErr.Error()
+	partialResp := ""
+	if resultMA != nil {
+		partialResp = resultMA.Response
+	}
+	isCancelled := errors.Is(context.Cause(baseCtx), ErrTaskCancelled) ||
+		errors.Is(runErr, context.Canceled) ||
+		strings.Contains(strings.ToLower(errStr), "context canceled") ||
+		strings.Contains(strings.ToLower(errStr), "context cancelled") ||
+		(partialResp != "" && (strings.Contains(partialResp, "任务已被取消") || strings.Contains(partialResp, "任务执行中断")))
+	isTimeout := errors.Is(runErr, context.DeadlineExceeded) || errors.Is(context.Cause(taskCtx), context.DeadlineExceeded)
+
+	if isTimeout {
+		*finishStatus = "timeout"
+	} else if isCancelled {
+		*finishStatus = "cancelled"
+	} else {
+		*finishStatus = "failed"
+	}
+
+	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。"
+		if partialResp != "" && (strings.Contains(partialResp, "任务已被取消") || strings.Contains(partialResp, "任务执行中断")) {
+			cancelMsg = partialResp
+		}
+		if assistantMessageID != "" {
+			if updateErr := h.appendAssistantMessageNotice(assistantMessageID, cancelMsg); updateErr != nil {
+				h.logger.Warn("更新取消后的助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(updateErr))
+			}
+			if err := h.db.AddProcessDetail(assistantMessageID, conversationID, "cancelled", cancelMsg, nil); err != nil {
+				h.logger.Warn("保存取消详情失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(err))
+			}
+		} else if _, errMsg := h.db.AddMessage(conversationID, "assistant", cancelMsg, nil); errMsg != nil {
+			h.logger.Warn("保存取消消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(errMsg))
+		}
+		h.batchTaskManager.UpdateTaskStatusWithConversationID(queueID, task.ID, BatchTaskStatusCancelled, cancelMsg, "", conversationID)
+		return
+	}
+
+	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()
+	if assistantMessageID != "" {
+		if _, updateErr := h.db.Exec(
+			"UPDATE messages SET content = ?, updated_at = ? WHERE id = ?",
+			errorMsg,
+			time.Now(), assistantMessageID,
+		); updateErr != nil {
+			h.logger.Warn("更新失败后的助手消息失败", zap.String("queueId", queueID), zap.String("taskId", task.ID), zap.Error(updateErr))
+		}
+		if err := h.db.AddProcessDetail(assistantMessageID, conversationID, "error", errorMsg, nil); err != nil {
+			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, BatchTaskStatusFailed, "", runErr.Error())
+}

internal/handler/batch_task_manager.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"crypto/rand"
 	"encoding/hex"
+	"errors"
 	"fmt"
 	"sort"
 	"strings"
@@ -17,6 +18,15 @@ import (
 	"go.uber.org/zap"
 )
 
+var (
+	// ErrBatchQueueNotFound 队列不存在或已从内存卸载。
+	ErrBatchQueueNotFound = errors.New("batch queue not found")
+	// ErrBatchQueueExecutorActive executeBatchQueue 协程仍在收尾，禁止删除。
+	ErrBatchQueueExecutorActive = errors.New("batch queue executor is still active")
+	// ErrBatchQueueStillRunning 队列状态仍为 running（无活跃执行器时的兜底保护）。
+	ErrBatchQueueStillRunning = errors.New("batch queue is still running")
+)
+
 // 批量任务状态常量
 const (
 	BatchQueueStatusPending   = "pending"
@@ -39,6 +49,12 @@ const (
 
 	// MaxBatchQueueRoleLen 角色名最大长度
 	MaxBatchQueueRoleLen = 100
+
+	// DefaultBatchQueueConcurrency 批量队列默认并发数（串行）
+	DefaultBatchQueueConcurrency = 1
+
+	// MaxBatchQueueConcurrency 批量队列最大并发数
+	MaxBatchQueueConcurrency = 8
 )
 
 // BatchTask 批量任务项
@@ -67,6 +83,7 @@ type BatchTaskQueue struct {
 	LastScheduleError     string       `json:"lastScheduleError,omitempty"`
 	LastRunError          string       `json:"lastRunError,omitempty"`
 	ProjectID             string       `json:"projectId,omitempty"`
+	Concurrency           int          `json:"concurrency"` // 同时执行的子任务数，默认 1
 	Tasks                 []*BatchTask `json:"tasks"`
 	Status                string       `json:"status"` // pending, running, paused, completed, cancelled
 	CreatedAt             time.Time    `json:"createdAt"`
@@ -80,8 +97,9 @@ type BatchTaskManager struct {
 	db             *database.DB
 	logger         *zap.Logger
 	queues         map[string]*BatchTaskQueue
-	taskCancels    map[string]context.CancelFunc // 存储每个队列当前任务的取消函数
+	taskCancels    map[string]map[string]context.CancelFunc // queueID -> taskID -> 取消函数
 	singleRunTasks map[string]string             // queueID -> taskID，单条执行完成后暂停队列
+	queueExecutors map[string]struct{}           // executeBatchQueue 协程活跃标记（与队列 status 解耦）
 	mu             sync.RWMutex
 }
 
@@ -93,11 +111,56 @@ func NewBatchTaskManager(logger *zap.Logger) *BatchTaskManager {
 	return &BatchTaskManager{
 		logger:         logger,
 		queues:         make(map[string]*BatchTaskQueue),
-		taskCancels:    make(map[string]context.CancelFunc),
+		taskCancels:    make(map[string]map[string]context.CancelFunc),
 		singleRunTasks: make(map[string]string),
+		queueExecutors: make(map[string]struct{}),
 	}
 }
 
+// batchQueueExecutionShouldStop 判断 executeBatchQueue 主循环是否应退出。
+func batchQueueExecutionShouldStop(queue *BatchTaskQueue, exists bool) bool {
+	if !exists || queue == nil {
+		return true
+	}
+	switch queue.Status {
+	case BatchQueueStatusCancelled, BatchQueueStatusCompleted, BatchQueueStatusPaused:
+		return true
+	default:
+		return false
+	}
+}
+
+// TryMarkQueueExecutor 标记队列执行协程已启动；若已有执行协程则返回 false。
+func (m *BatchTaskManager) TryMarkQueueExecutor(queueID string) bool {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	if _, exists := m.queueExecutors[queueID]; exists {
+		return false
+	}
+	m.queueExecutors[queueID] = struct{}{}
+	return true
+}
+
+// UnmarkQueueExecutor 清除队列执行协程标记（executeBatchQueue defer 调用）。
+func (m *BatchTaskManager) UnmarkQueueExecutor(queueID string) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	delete(m.queueExecutors, queueID)
+}
+
+// ForceUnmarkQueueExecutor 强制清除执行协程标记（暂停态单条重跑等场景回收陈旧槽位）。
+func (m *BatchTaskManager) ForceUnmarkQueueExecutor(queueID string) {
+	m.UnmarkQueueExecutor(queueID)
+}
+
+// IsQueueExecutorActive 队列 executeBatchQueue 协程是否仍在运行。
+func (m *BatchTaskManager) IsQueueExecutorActive(queueID string) bool {
+	m.mu.RLock()
+	defer m.mu.RUnlock()
+	_, ok := m.queueExecutors[queueID]
+	return ok
+}
+
 // SetDB 设置数据库连接
 func (m *BatchTaskManager) SetDB(db *database.DB) {
 	m.mu.Lock()
@@ -105,10 +168,22 @@ func (m *BatchTaskManager) SetDB(db *database.DB) {
 	m.db = db
 }
 
+// normalizeBatchQueueConcurrency 规范化队列并发数。
+func normalizeBatchQueueConcurrency(n int) int {
+	if n < 1 {
+		return DefaultBatchQueueConcurrency
+	}
+	if n > MaxBatchQueueConcurrency {
+		return MaxBatchQueueConcurrency
+	}
+	return n
+}
+
 // CreateBatchQueue 创建批量任务队列
 func (m *BatchTaskManager) CreateBatchQueue(
 	title, role, agentMode, scheduleMode, cronExpr, projectID string,
 	nextRunAt *time.Time,
+	concurrency int,
 	tasks []string,
 ) (*BatchTaskQueue, error) {
 	// 输入校验
@@ -136,6 +211,7 @@ func (m *BatchTaskManager) CreateBatchQueue(
 		CronExpr:        strings.TrimSpace(cronExpr),
 		NextRunAt:       nextRunAt,
 		ScheduleEnabled: true,
+		Concurrency:     normalizeBatchQueueConcurrency(concurrency),
 		Tasks:           make([]*BatchTask, 0, len(tasks)),
 		Status:          BatchQueueStatusPending,
 		CreatedAt:       time.Now(),
@@ -177,6 +253,7 @@ func (m *BatchTaskManager) CreateBatchQueue(
 			queue.CronExpr,
 			queue.NextRunAt,
 			queue.ProjectID,
+			queue.Concurrency,
 			dbTasks,
 		); err != nil {
 			m.logger.Warn("batch queue DB create failed", zap.String("queueId", queueID), zap.Error(err))
@@ -272,6 +349,7 @@ func (m *BatchTaskManager) loadQueueFromDB(queueID string) *BatchTaskQueue {
 	if queueRow.ProjectID.Valid {
 		queue.ProjectID = strings.TrimSpace(queueRow.ProjectID.String)
 	}
+	queue.Concurrency = batchQueueConcurrencyFromRow(queueRow)
 	if queueRow.StartedAt.Valid {
 		queue.StartedAt = &queueRow.StartedAt.Time
 	}
@@ -511,6 +589,7 @@ func (m *BatchTaskManager) LoadFromDB() error {
 		if queueRow.ProjectID.Valid {
 			queue.ProjectID = strings.TrimSpace(queueRow.ProjectID.String)
 		}
+		queue.Concurrency = batchQueueConcurrencyFromRow(queueRow)
 		if queueRow.StartedAt.Valid {
 			queue.StartedAt = &queueRow.StartedAt.Time
 		}
@@ -651,8 +730,16 @@ func (m *BatchTaskManager) UpdateQueueSchedule(queueID, scheduleMode, cronExpr s
 	}
 }
 
-// UpdateQueueMetadata 更新队列标题、角色和代理模式（非 running 时可用）
+// batchQueueConcurrencyFromRow 从数据库行读取并发数（缺省为 1）。
-func (m *BatchTaskManager) UpdateQueueMetadata(queueID, title, role, agentMode string) error {
+func batchQueueConcurrencyFromRow(row *database.BatchTaskQueueRow) int {
+	if row == nil || !row.Concurrency.Valid {
+		return DefaultBatchQueueConcurrency
+	}
+	return normalizeBatchQueueConcurrency(int(row.Concurrency.Int64))
+}
+
+// UpdateQueueMetadata 更新队列标题、角色、代理模式和并发数（非 running 时可用）
+func (m *BatchTaskManager) UpdateQueueMetadata(queueID, title, role, agentMode string, concurrency *int) error {
 	if utf8.RuneCountInString(title) > MaxBatchQueueTitleLen {
 		return fmt.Errorf("标题不能超过 %d 个字符", MaxBatchQueueTitleLen)
 	}
@@ -680,9 +767,12 @@ func (m *BatchTaskManager) UpdateQueueMetadata(queueID, title, role, agentMode s
 	queue.Title = title
 	queue.Role = role
 	queue.AgentMode = agentMode
+	if concurrency != nil {
+		queue.Concurrency = normalizeBatchQueueConcurrency(*concurrency)
+	}
 
 	if m.db != nil {
-		if err := m.db.UpdateBatchQueueMetadata(queueID, title, role, agentMode); err != nil {
+		if err := m.db.UpdateBatchQueueMetadata(queueID, title, role, agentMode, queue.Concurrency); err != nil {
 			m.logger.Warn("batch queue DB metadata update failed", zap.String("queueId", queueID), zap.Error(err))
 		}
 	}
@@ -868,7 +958,6 @@ func (m *BatchTaskManager) AddTaskToQueue(queueID, message string) (*BatchTask,
 
 // PrepareSingleTaskRun 准备单条执行：重置目标任务（若已有结果）并定位队列索引
 func (m *BatchTaskManager) PrepareSingleTaskRun(queueID, taskID string) error {
-	var cancelFunc context.CancelFunc
 	var siblingRunningIDs []string
 
 	m.mu.Lock()
@@ -898,11 +987,9 @@ func (m *BatchTaskManager) PrepareSingleTaskRun(queueID, taskID string) error {
 	}
 
 	// 暂停态：中止在途子任务并收口仍标记 running 的其它子任务，以便单条执行非冲突项
+	var cancelFuncs []context.CancelFunc
 	if queue.Status == BatchQueueStatusPaused {
-		if c, ok := m.taskCancels[queueID]; ok {
+		cancelFuncs = m.drainTaskCancelsLocked(queueID)
-			cancelFunc = c
-			delete(m.taskCancels, queueID)
-		}
 		for _, t := range queue.Tasks {
 			if t != nil && t.ID != taskID && t.Status == BatchTaskStatusRunning {
 				siblingRunningIDs = append(siblingRunningIDs, t.ID)
@@ -914,8 +1001,10 @@ func (m *BatchTaskManager) PrepareSingleTaskRun(queueID, taskID string) error {
 	resumeQueue := queue.Status == BatchQueueStatusCompleted || queue.Status == BatchQueueStatusCancelled
 	m.mu.Unlock()
 
-	if cancelFunc != nil {
+	for _, c := range cancelFuncs {
-		cancelFunc()
+		if c != nil {
+			c()
+		}
 	}
 	const staleRunMsg = "为单条执行其它任务，已中止"
 	for _, sid := range siblingRunningIDs {
@@ -1089,7 +1178,90 @@ func queueAllowsSingleTaskRunLocked(queue *BatchTaskQueue, task *BatchTask) bool
 	}
 }
 
-// GetNextTask 获取下一个待执行的任务
+// ClaimNextPendingTask 原子领取下一个待执行子任务（并发 worker 安全）。
+func (m *BatchTaskManager) ClaimNextPendingTask(queueID string) (*BatchTask, bool) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	queue, exists := m.queues[queueID]
+	if !exists || queue == nil {
+		return nil, false
+	}
+	if queue.Status == BatchQueueStatusCancelled || queue.Status == BatchQueueStatusCompleted || queue.Status == BatchQueueStatusPaused {
+		return nil, false
+	}
+
+	onlyTaskID := ""
+	if m.singleRunTasks != nil {
+		onlyTaskID = m.singleRunTasks[queueID]
+	}
+
+	for i, task := range queue.Tasks {
+		if task == nil || task.Status != BatchTaskStatusPending {
+			continue
+		}
+		if onlyTaskID != "" && task.ID != onlyTaskID {
+			continue
+		}
+		task.Status = BatchTaskStatusRunning
+		queue.CurrentIndex = i
+		return task, true
+	}
+	return nil, false
+}
+
+// HasRunningTasks 队列是否仍有 running 状态的子任务。
+func (m *BatchTaskManager) HasRunningTasks(queueID string) bool {
+	m.mu.RLock()
+	defer m.mu.RUnlock()
+	queue, exists := m.queues[queueID]
+	if !exists || queue == nil {
+		return false
+	}
+	for _, task := range queue.Tasks {
+		if task != nil && task.Status == BatchTaskStatusRunning {
+			return true
+		}
+	}
+	return false
+}
+
+// HasPendingOrRunningTasks 队列是否仍有未完成的子任务。
+func (m *BatchTaskManager) HasPendingOrRunningTasks(queueID string) bool {
+	m.mu.RLock()
+	defer m.mu.RUnlock()
+	queue, exists := m.queues[queueID]
+	if !exists || queue == nil {
+		return false
+	}
+	for _, task := range queue.Tasks {
+		if task == nil {
+			continue
+		}
+		if task.Status == BatchTaskStatusPending || task.Status == BatchTaskStatusRunning {
+			return true
+		}
+	}
+	return false
+}
+
+// drainTaskCancelsLocked 取出并清空队列下所有子任务取消函数（调用方须已持 m.mu）。
+func (m *BatchTaskManager) drainTaskCancelsLocked(queueID string) []context.CancelFunc {
+	taskMap, ok := m.taskCancels[queueID]
+	if !ok || len(taskMap) == 0 {
+		return nil
+	}
+	cancels := make([]context.CancelFunc, 0, len(taskMap))
+	for _, c := range taskMap {
+		if c != nil {
+			cancels = append(cancels, c)
+		}
+	}
+	delete(m.taskCancels, queueID)
+	return cancels
+}
+
+// GetNextTask 获取下一个待执行的任务（串行兼容，优先使用 ClaimNextPendingTask）
 func (m *BatchTaskManager) GetNextTask(queueID string) (*BatchTask, bool) {
 	m.mu.Lock()
 	defer m.mu.Unlock()
@@ -1130,20 +1302,28 @@ func (m *BatchTaskManager) MoveToNextTask(queueID string) {
 	}
 }
 
-// SetTaskCancel 设置当前任务的取消函数
+// SetTaskCancel 设置子任务的取消函数
-func (m *BatchTaskManager) SetTaskCancel(queueID string, cancel context.CancelFunc) {
+func (m *BatchTaskManager) SetTaskCancel(queueID, taskID string, cancel context.CancelFunc) {
 	m.mu.Lock()
 	defer m.mu.Unlock()
-	if cancel != nil {
+	if cancel == nil {
-		m.taskCancels[queueID] = cancel
+		if taskMap, ok := m.taskCancels[queueID]; ok {
-	} else {
+			delete(taskMap, taskID)
+			if len(taskMap) == 0 {
 				delete(m.taskCancels, queueID)
 			}
 		}
+		return
+	}
+	if m.taskCancels[queueID] == nil {
+		m.taskCancels[queueID] = make(map[string]context.CancelFunc)
+	}
+	m.taskCancels[queueID][taskID] = cancel
+}
 
 // PauseQueue 暂停队列
 func (m *BatchTaskManager) PauseQueue(queueID string) bool {
-	var cancelFunc context.CancelFunc
+	var cancelFuncs []context.CancelFunc
 
 	m.mu.Lock()
 	queue, exists := m.queues[queueID]
@@ -1168,17 +1348,11 @@ func (m *BatchTaskManager) PauseQueue(queueID string) bool {
 	}
 
 	queue.Status = BatchQueueStatusPaused
-
+	cancelFuncs = m.drainTaskCancelsLocked(queueID)
-	// 取消当前正在执行的任务（通过取消context）
-	if cancel, ok := m.taskCancels[queueID]; ok {
-		cancelFunc = cancel
-		delete(m.taskCancels, queueID)
-	}
 	m.mu.Unlock()
 
-	// 释放锁后执行取消回调（cancel 可能阻塞，不应持锁）
+	for _, c := range cancelFuncs {
-	if cancelFunc != nil {
+		c()
-		cancelFunc()
 	}
 
 	return true
@@ -1187,7 +1361,7 @@ func (m *BatchTaskManager) PauseQueue(queueID string) bool {
 // CancelQueue 取消队列（保留此方法以保持向后兼容，但建议使用PauseQueue）
 func (m *BatchTaskManager) CancelQueue(queueID string) bool {
 	now := time.Now()
-	var cancelFunc context.CancelFunc
+	var cancelFuncs []context.CancelFunc
 
 	m.mu.Lock()
 	queue, exists := m.queues[queueID]
@@ -1228,34 +1402,33 @@ func (m *BatchTaskManager) CancelQueue(queueID string) bool {
 		}
 	}
 
-	// 取消当前正在执行的任务
+	cancelFuncs = m.drainTaskCancelsLocked(queueID)
-	if cancel, ok := m.taskCancels[queueID]; ok {
-		cancelFunc = cancel
-		delete(m.taskCancels, queueID)
-	}
 	m.mu.Unlock()
 
-	// 释放锁后执行取消回调（cancel 可能阻塞，不应持锁）
+	for _, c := range cancelFuncs {
-	if cancelFunc != nil {
+		c()
-		cancelFunc()
 	}
 
 	return true
 }
 
-// DeleteQueue 删除队列（运行中的队列不允许删除）
+// DeleteQueue 删除队列。执行协程活跃或 status 为 running 时拒绝删除，避免 executeBatchQueue 空指针 panic。
-func (m *BatchTaskManager) DeleteQueue(queueID string) bool {
+func (m *BatchTaskManager) DeleteQueue(queueID string) error {
 	m.mu.Lock()
 	defer m.mu.Unlock()
 
 	queue, exists := m.queues[queueID]
 	if !exists {
-		return false
+		return ErrBatchQueueNotFound
+	}
+
+	if _, exec := m.queueExecutors[queueID]; exec {
+		return ErrBatchQueueExecutorActive
 	}
 
 	// 运行中的队列不允许删除，防止孤儿协程和数据丢失
 	if queue.Status == BatchQueueStatusRunning {
-		return false
+		return ErrBatchQueueStillRunning
 	}
 
 	// 清理取消函数
@@ -1269,7 +1442,7 @@ func (m *BatchTaskManager) DeleteQueue(queueID string) bool {
 	}
 
 	delete(m.queues, queueID)
-	return true
+	return nil
 }
 
 // generateShortID 生成短ID

internal/handler/batch_task_manager_test.go

@@ -0,0 +1,121 @@
+package handler
+
+import (
+	"errors"
+	"testing"
+
+	"go.uber.org/zap"
+)
+
+func TestNormalizeBatchQueueConcurrency(t *testing.T) {
+	if got := normalizeBatchQueueConcurrency(0); got != DefaultBatchQueueConcurrency {
+		t.Fatalf("expected default %d, got %d", DefaultBatchQueueConcurrency, got)
+	}
+	if got := normalizeBatchQueueConcurrency(99); got != MaxBatchQueueConcurrency {
+		t.Fatalf("expected max %d, got %d", MaxBatchQueueConcurrency, got)
+	}
+}
+
+func TestClaimNextPendingTaskParallel(t *testing.T) {
+	m := NewBatchTaskManager(zap.NewNop())
+	queue, err := m.CreateBatchQueue("test", "", "eino_single", "manual", "", "", nil, 3, []string{"a", "b", "c"})
+	if err != nil {
+		t.Fatalf("CreateBatchQueue: %v", err)
+	}
+	m.UpdateQueueStatus(queue.ID, BatchQueueStatusRunning)
+
+	t1, ok1 := m.ClaimNextPendingTask(queue.ID)
+	t2, ok2 := m.ClaimNextPendingTask(queue.ID)
+	if !ok1 || !ok2 || t1.ID == t2.ID {
+		t.Fatalf("expected two distinct claims, got ok1=%v ok2=%v t1=%v t2=%v", ok1, ok2, t1, t2)
+	}
+	if t1.Status != BatchTaskStatusRunning || t2.Status != BatchTaskStatusRunning {
+		t.Fatalf("claimed tasks should be running")
+	}
+	t3, ok3 := m.ClaimNextPendingTask(queue.ID)
+	if !ok3 {
+		t.Fatal("expected third claim")
+	}
+	_, ok4 := m.ClaimNextPendingTask(queue.ID)
+	if ok4 {
+		t.Fatal("expected no fourth pending task")
+	}
+	_ = t3
+}
+
+func TestBatchQueueExecutionShouldStop(t *testing.T) {
+	t.Parallel()
+	if !batchQueueExecutionShouldStop(nil, false) {
+		t.Fatal("expected stop when queue missing")
+	}
+	if !batchQueueExecutionShouldStop(nil, true) {
+		t.Fatal("expected stop when queue is nil but exists=true")
+	}
+	q := &BatchTaskQueue{Status: BatchQueueStatusRunning}
+	if batchQueueExecutionShouldStop(q, true) {
+		t.Fatal("expected continue when running")
+	}
+	q.Status = BatchQueueStatusCancelled
+	if !batchQueueExecutionShouldStop(q, true) {
+		t.Fatal("expected stop when cancelled")
+	}
+}
+
+func TestDeleteQueueBlockedWhileExecutorActive(t *testing.T) {
+	t.Parallel()
+	m := NewBatchTaskManager(zap.NewNop())
+	queue, err := m.CreateBatchQueue("test", "", "eino_single", "manual", "", "", nil, 1, []string{"hello"})
+	if err != nil {
+		t.Fatalf("CreateBatchQueue: %v", err)
+	}
+	if !m.TryMarkQueueExecutor(queue.ID) {
+		t.Fatal("expected to mark executor")
+	}
+	m.UpdateQueueStatus(queue.ID, BatchQueueStatusCancelled)
+
+	err = m.DeleteQueue(queue.ID)
+	if !errors.Is(err, ErrBatchQueueExecutorActive) {
+		t.Fatalf("expected ErrBatchQueueExecutorActive, got %v", err)
+	}
+	if _, ok := m.GetBatchQueue(queue.ID); !ok {
+		t.Fatal("queue should still exist while executor active")
+	}
+
+	m.UnmarkQueueExecutor(queue.ID)
+	if err := m.DeleteQueue(queue.ID); err != nil {
+		t.Fatalf("expected delete after executor unmarked, got %v", err)
+	}
+	if _, ok := m.GetBatchQueue(queue.ID); ok {
+		t.Fatal("queue should be deleted")
+	}
+}
+
+func TestDeleteQueueBlockedWhileRunning(t *testing.T) {
+	t.Parallel()
+	m := NewBatchTaskManager(zap.NewNop())
+	queue, err := m.CreateBatchQueue("test", "", "eino_single", "manual", "", "", nil, 1, []string{"hello"})
+	if err != nil {
+		t.Fatalf("CreateBatchQueue: %v", err)
+	}
+	m.UpdateQueueStatus(queue.ID, BatchQueueStatusRunning)
+
+	err = m.DeleteQueue(queue.ID)
+	if !errors.Is(err, ErrBatchQueueStillRunning) {
+		t.Fatalf("expected ErrBatchQueueStillRunning, got %v", err)
+	}
+}
+
+func TestTryMarkQueueExecutorDedupes(t *testing.T) {
+	t.Parallel()
+	m := NewBatchTaskManager(zap.NewNop())
+	if !m.TryMarkQueueExecutor("q-1") {
+		t.Fatal("first mark should succeed")
+	}
+	if m.TryMarkQueueExecutor("q-1") {
+		t.Fatal("second mark should fail")
+	}
+	m.UnmarkQueueExecutor("q-1")
+	if !m.TryMarkQueueExecutor("q-1") {
+		t.Fatal("mark after unmark should succeed")
+	}
+}

internal/handler/batch_task_mcp.go

@@ -3,6 +3,7 @@ package handler
 import (
 	"context"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"strconv"
 	"strings"
@@ -181,6 +182,10 @@ func RegisterBatchTaskMCPTools(mcpServer *mcp.Server, h *AgentHandler, logger *z
 					"type":        "string",
 					"description": "队列内子对话绑定的项目 ID（可选，未指定时使用 config.project.default_project_id）",
 				},
+				"concurrency": map[string]interface{}{
+					"type":        "integer",
+					"description": "同时执行的子任务数，默认 1（串行），最大 8。含扫描类工具时建议 1-2。",
+				},
 			},
 		},
 	}, func(ctx context.Context, args map[string]interface{}) (*mcp.ToolResult, error) {
@@ -210,7 +215,8 @@ func RegisterBatchTaskMCPTools(mcpServer *mcp.Server, h *AgentHandler, logger *z
 			executeNow = false
 		}
 		projectID := strings.TrimSpace(mcpArgString(args, "project_id"))
-		queue, createErr := h.batchTaskManager.CreateBatchQueue(title, role, agentMode, scheduleMode, cronExpr, projectID, nextRunAt, tasks)
+		concurrency := int(mcpArgFloat(args, "concurrency"))
+		queue, createErr := h.batchTaskManager.CreateBatchQueue(title, role, agentMode, scheduleMode, cronExpr, projectID, nextRunAt, concurrency, tasks)
 		if createErr != nil {
 			return batchMCPTextResult("创建队列失败: "+createErr.Error(), true), nil
 		}
@@ -365,8 +371,17 @@ func RegisterBatchTaskMCPTools(mcpServer *mcp.Server, h *AgentHandler, logger *z
 		if qid == "" {
 			return batchMCPTextResult("queue_id 不能为空", true), nil
 		}
-		if !h.batchTaskManager.DeleteQueue(qid) {
+		if err := h.batchTaskManager.DeleteQueue(qid); err != nil {
+			switch {
+			case errors.Is(err, ErrBatchQueueNotFound):
 				return batchMCPTextResult("删除失败：队列不存在", true), nil
+			case errors.Is(err, ErrBatchQueueExecutorActive):
+				return batchMCPTextResult("删除失败：队列执行器仍在运行，请稍后再试", true), nil
+			case errors.Is(err, ErrBatchQueueStillRunning):
+				return batchMCPTextResult("删除失败：队列正在运行中", true), nil
+			default:
+				return batchMCPTextResult("删除失败："+err.Error(), true), nil
+			}
 		}
 		logger.Info("MCP batch_task_delete", zap.String("queueId", qid))
 		return batchMCPTextResult("队列已删除。", false), nil
@@ -397,6 +412,10 @@ func RegisterBatchTaskMCPTools(mcpServer *mcp.Server, h *AgentHandler, logger *z
 					"description": "代理模式：eino_single、deep、plan_execute、supervisor",
 					"enum":        []string{"eino_single", "deep", "plan_execute", "supervisor"},
 				},
+				"concurrency": map[string]interface{}{
+					"type":        "integer",
+					"description": "同时执行的子任务数，默认 1，最大 8",
+				},
 			},
 			"required": []string{"queue_id"},
 		},
@@ -408,7 +427,12 @@ func RegisterBatchTaskMCPTools(mcpServer *mcp.Server, h *AgentHandler, logger *z
 		title := mcpArgString(args, "title")
 		role := mcpArgString(args, "role")
 		agentMode := mcpArgString(args, "agent_mode")
-		if err := h.batchTaskManager.UpdateQueueMetadata(qid, title, role, agentMode); err != nil {
+		var concurrency *int
+		if raw, ok := args["concurrency"]; ok && raw != nil {
+			v := int(mcpArgFloat(args, "concurrency"))
+			concurrency = &v
+		}
+		if err := h.batchTaskManager.UpdateQueueMetadata(qid, title, role, agentMode, concurrency); err != nil {
 			return batchMCPTextResult(err.Error(), true), nil
 		}
 		updated, _ := h.batchTaskManager.GetBatchQueue(qid)
@@ -652,6 +676,7 @@ type batchTaskQueueMCPListItem struct {
 	StartedAt             *time.Time                `json:"startedAt,omitempty"`
 	CompletedAt           *time.Time                `json:"completedAt,omitempty"`
 	CurrentIndex          int                       `json:"currentIndex"`
+	Concurrency           int                       `json:"concurrency"`
 	TaskTotal             int                       `json:"task_total"`
 	TaskCounts            map[string]int            `json:"task_counts"`
 	Tasks                 []batchTaskMCPListSummary `json:"tasks"`
@@ -715,6 +740,7 @@ func toBatchTaskQueueMCPListItem(q *BatchTaskQueue) batchTaskQueueMCPListItem {
 		StartedAt:             q.StartedAt,
 		CompletedAt:           q.CompletedAt,
 		CurrentIndex:          q.CurrentIndex,
+		Concurrency:           q.Concurrency,
 		TaskTotal:             len(tasks),
 		TaskCounts:            counts,
 		Tasks:                 tasks,

internal/handler/conversation.go

@@ -12,11 +12,17 @@ import (
 	"go.uber.org/zap"
 )
 
+// ConversationTaskStopper cancels in-flight agent work when a conversation is removed.
+type ConversationTaskStopper interface {
+	CancelRunningTaskForConversation(conversationID string)
+}
+
 // ConversationHandler 对话处理器
 type ConversationHandler struct {
 	db          *database.DB
 	logger      *zap.Logger
 	audit       *audit.Service
+	taskStopper ConversationTaskStopper
 }
 
 // SetAudit wires platform audit logging.
@@ -24,6 +30,11 @@ func (h *ConversationHandler) SetAudit(s *audit.Service) {
 	h.audit = s
 }
 
+// SetTaskStopper wires cancellation of in-flight agent tasks on conversation delete.
+func (h *ConversationHandler) SetTaskStopper(stopper ConversationTaskStopper) {
+	h.taskStopper = stopper
+}
+
 // NewConversationHandler 创建新的对话处理器
 func NewConversationHandler(db *database.DB, logger *zap.Logger) *ConversationHandler {
 	return &ConversationHandler{
@@ -165,6 +176,9 @@ func (h *ConversationHandler) GetConversation(c *gin.Context) {
 }
 
 // GetMessageProcessDetails 获取指定消息的过程详情（按需加载）
+// 查询参数：
+//   - summary=1：仅返回摘要（total / iterationCount / maxIteration）
+//   - limit + offset：分页返回 processDetails（未指定 limit 时保持全量兼容）
 func (h *ConversationHandler) GetMessageProcessDetails(c *gin.Context) {
 	messageID := c.Param("id")
 	if messageID == "" {
@@ -172,6 +186,51 @@ func (h *ConversationHandler) GetMessageProcessDetails(c *gin.Context) {
 		return
 	}
 
+	summaryStr := strings.TrimSpace(c.Query("summary"))
+	if summaryStr == "1" || strings.EqualFold(summaryStr, "true") || strings.EqualFold(summaryStr, "yes") {
+		summary, err := h.db.GetProcessDetailsSummary(messageID)
+		if err != nil {
+			h.logger.Error("获取过程详情摘要失败", zap.Error(err))
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		c.JSON(http.StatusOK, gin.H{"summary": summary})
+		return
+	}
+
+	limitStr := strings.TrimSpace(c.Query("limit"))
+	if limitStr != "" {
+		limit, err := strconv.Atoi(limitStr)
+		if err != nil || limit <= 0 {
+			c.JSON(http.StatusBadRequest, gin.H{"error": "invalid limit"})
+			return
+		}
+		if limit > 500 {
+			limit = 500
+		}
+		offset, _ := strconv.Atoi(strings.TrimSpace(c.Query("offset")))
+		if offset < 0 {
+			offset = 0
+		}
+
+		details, total, err := h.db.GetProcessDetailsPage(messageID, limit, offset)
+		if err != nil {
+			h.logger.Error("分页获取过程详情失败", zap.Error(err))
+			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+			return
+		}
+		details = database.DedupeConsecutiveProcessDetails(details)
+		out := processDetailsToJSON(h.logger, details)
+		c.JSON(http.StatusOK, gin.H{
+			"processDetails": out,
+			"total":          total,
+			"offset":         offset,
+			"limit":          limit,
+			"hasMore":        offset+len(out) < total,
+		})
+		return
+	}
+
 	details, err := h.db.GetProcessDetails(messageID)
 	if err != nil {
 		h.logger.Error("获取过程详情失败", zap.Error(err))
@@ -180,14 +239,17 @@ func (h *ConversationHandler) GetMessageProcessDetails(c *gin.Context) {
 	}
 
 	details = database.DedupeConsecutiveProcessDetails(details)
+	out := processDetailsToJSON(h.logger, details)
+	c.JSON(http.StatusOK, gin.H{"processDetails": out, "total": len(out)})
+}
 
-	// 转换为前端期望的 JSON 结构（与 GetConversation 中 processDetails 结构一致）
+func processDetailsToJSON(logger *zap.Logger, details []database.ProcessDetail) []map[string]interface{} {
 	out := make([]map[string]interface{}, 0, len(details))
 	for _, d := range details {
 		var data interface{}
 		if d.Data != "" {
 			if err := json.Unmarshal([]byte(d.Data), &data); err != nil {
-				h.logger.Warn("解析过程详情数据失败", zap.Error(err))
+				logger.Warn("解析过程详情数据失败", zap.Error(err))
 			}
 		}
 		out = append(out, map[string]interface{}{
@@ -200,8 +262,7 @@ func (h *ConversationHandler) GetMessageProcessDetails(c *gin.Context) {
 			"createdAt":      d.CreatedAt,
 		})
 	}
-
+	return out
-	c.JSON(http.StatusOK, gin.H{"processDetails": out})
 }
 
 // UpdateConversationRequest 更新对话请求
@@ -245,6 +306,10 @@ func (h *ConversationHandler) UpdateConversation(c *gin.Context) {
 func (h *ConversationHandler) DeleteConversation(c *gin.Context) {
 	id := c.Param("id")
 
+	if h.taskStopper != nil {
+		h.taskStopper.CancelRunningTaskForConversation(id)
+	}
+
 	if err := h.db.DeleteConversation(id); err != nil {
 		h.logger.Error("删除对话失败", zap.Error(err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})

internal/handler/conversation_delete_task_test.go

@@ -0,0 +1,30 @@
+package handler
+
+import (
+	"context"
+	"testing"
+	"time"
+
+	"go.uber.org/zap"
+)
+
+func TestConversationHandlerDeleteConversationCancelsRunningTask(t *testing.T) {
+	tm := NewAgentTaskManager()
+	ctx, cancel := context.WithCancelCause(context.Background())
+	_, err := tm.StartTask("conv-1", "hello", cancel)
+	if err != nil {
+		t.Fatalf("StartTask: %v", err)
+	}
+
+	h := &AgentHandler{tasks: tm, logger: zap.NewNop()}
+	h.CancelRunningTaskForConversation("conv-1")
+
+	select {
+	case <-ctx.Done():
+	case <-time.After(2 * time.Second):
+		t.Fatal("task context was not cancelled")
+	}
+	if cause := context.Cause(ctx); cause != ErrTaskCancelled {
+		t.Fatalf("expected ErrTaskCancelled, got %v", cause)
+	}
+}

internal/handler/eino_resume_segment.go

@@ -2,31 +2,11 @@ package handler
 
 import (
 	"context"
-	"errors"
-	"fmt"
-	"strings"
-	"time"
 
 	"cyberstrike-ai/internal/agent"
 	"cyberstrike-ai/internal/multiagent"
-
-	"go.uber.org/zap"
 )
 
-func (h *AgentHandler) einoRunRetryMaxAttempts() int {
-	if h.config != nil {
-		return multiagent.RunRetryMaxAttemptsFromConfig(&h.config.MultiAgent.EinoMiddleware)
-	}
-	return multiagent.RunRetryMaxAttemptsFromConfig(nil)
-}
-
-func (h *AgentHandler) einoRunRetryMaxBackoffSec() int {
-	if h.config != nil && h.config.MultiAgent.EinoMiddleware.RunRetryMaxBackoffSec > 0 {
-		return h.config.MultiAgent.EinoMiddleware.RunRetryMaxBackoffSec
-	}
-	return 0
-}
-
 // applyEinoTraceResumeSegment 中断并继续：persist last_react_* → loadHistory，可选替换下一段 user 文案。
 func (h *AgentHandler) applyEinoTraceResumeSegment(
 	conversationID string,
@@ -45,136 +25,3 @@ func (h *AgentHandler) applyEinoTraceResumeSegment(
 		*curFinalMessage = segmentUserMessage
 	}
 }
-
-// applyEinoTransientRetrySegment 临时错误重试：恢复轨迹并保留本请求原始 user 文案（不注入续跑说明）。
-// segmentUserMessage 为本轮 HTTP 请求开始时用户发送的内容，避免因清空 finalMessage 而丢失「你好」等短句。
-func (h *AgentHandler) applyEinoTransientRetrySegment(
-	conversationID string,
-	result *multiagent.RunResult,
-	curHistory *[]agent.ChatMessage,
-	curFinalMessage *string,
-	segmentUserMessage string,
-) {
-	if shouldPersistEinoAgentTraceAfterRunError(context.Background()) {
-		h.persistEinoAgentTraceForResume(conversationID, result)
-	}
-	if hist, err := h.loadHistoryFromAgentTrace(conversationID); err == nil && len(hist) > 0 {
-		*curHistory = hist
-	}
-	if s := strings.TrimSpace(segmentUserMessage); s != "" {
-		*curFinalMessage = segmentUserMessage
-	}
-}
-
-// handleEinoTransientRetryContinue 在 SSE 任务循环内处理临时错误重试；返回 true 表示外层 for 应 continue。
-func (h *AgentHandler) handleEinoTransientRetryContinue(
-	baseCtx context.Context,
-	conversationID string,
-	result *multiagent.RunResult,
-	runErr error,
-	transientAttempts *int,
-	curHistory *[]agent.ChatMessage,
-	curFinalMessage *string,
-	segmentUserMessage string,
-	progressCallback func(eventType, message string, data interface{}),
-	sendProgress func(msg string, extra map[string]interface{}),
-) (handled bool, fatal error) {
-	if !errors.Is(runErr, multiagent.ErrTransientRetryContinue) {
-		return false, nil
-	}
-	maxAttempts := h.einoRunRetryMaxAttempts()
-	*transientAttempts++
-	if *transientAttempts > maxAttempts {
-		if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
-			h.persistEinoAgentTraceForResume(conversationID, result)
-		}
-		return false, errors.New("transient retry exhausted: " + runErr.Error())
-	}
-	attemptNo := *transientAttempts
-	backoff := multiagent.TransientRetryBackoff(attemptNo-1, h.einoRunRetryMaxBackoffSec())
-	if progressCallback != nil {
-		progressCallback("eino_run_retry", fmt.Sprintf("遇到临时错误，%d 秒后第 %d/%d 次重试…", int(backoff.Seconds()), attemptNo, maxAttempts), map[string]interface{}{
-			"conversationId": conversationID,
-			"source":         "eino",
-			"attempt":        attemptNo,
-			"maxAttempts":    maxAttempts,
-			"backoffSec":     int(backoff.Seconds()),
-		})
-	}
-	select {
-	case <-baseCtx.Done():
-		return false, context.Cause(baseCtx)
-	case <-time.After(backoff):
-	}
-	h.applyEinoTransientRetrySegment(conversationID, result, curHistory, curFinalMessage, segmentUserMessage)
-	if progressCallback != nil {
-		progressCallback("eino_run_retry", "已恢复上下文，正在重试…", map[string]interface{}{
-			"conversationId": conversationID,
-			"source":         "eino",
-			"attempt":        attemptNo,
-		})
-	}
-	if sendProgress != nil {
-		sendProgress("正在重试…", map[string]interface{}{
-			"conversationId": conversationID,
-			"source":         "transient_retry",
-		})
-	}
-	return true, nil
-}
-
-// handleEinoEmptyResponseContinue 在 SSE 任务循环内处理「正常结束但无助手正文」；返回 exhausted=true 时由外层按成功结束（保留占位文案）。
-// 与临时错误重试一致：仅恢复轨迹并保留本请求原始 user 文案，不向模型注入续跑说明。
-func (h *AgentHandler) handleEinoEmptyResponseContinue(
-	baseCtx context.Context,
-	conversationID string,
-	result *multiagent.RunResult,
-	runErr error,
-	emptyResponseAttempts *int,
-	curHistory *[]agent.ChatMessage,
-	curFinalMessage *string,
-	segmentUserMessage string,
-	progressCallback func(eventType, message string, data interface{}),
-	sendProgress func(msg string, extra map[string]interface{}),
-) (handled bool, exhausted bool) {
-	if !errors.Is(runErr, multiagent.ErrEmptyResponseContinue) {
-		return false, false
-	}
-	maxAttempts := h.einoRunRetryMaxAttempts()
-	*emptyResponseAttempts++
-	if *emptyResponseAttempts > maxAttempts {
-		if h.logger != nil {
-			h.logger.Warn("eino empty response auto resume exhausted",
-				zap.String("conversationId", conversationID),
-				zap.Int("maxAttempts", maxAttempts))
-		}
-		if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
-			h.persistEinoAgentTraceForResume(conversationID, result)
-		}
-		return false, true
-	}
-	attemptNo := *emptyResponseAttempts
-	if h.logger != nil {
-		h.logger.Info("eino empty response, auto resume from trace",
-			zap.String("conversationId", conversationID),
-			zap.Int("attempt", attemptNo),
-			zap.Int("maxAttempts", maxAttempts))
-	}
-	if progressCallback != nil {
-		progressCallback("eino_empty_response_continue", fmt.Sprintf("未捕获到助手正文，正在基于轨迹自动续跑（%d/%d）…", attemptNo, maxAttempts), map[string]interface{}{
-			"conversationId": conversationID,
-			"source":         "eino",
-			"attempt":        attemptNo,
-			"maxAttempts":    maxAttempts,
-			"resumeKind":     "trace_segment",
-		})
-	}
-	h.applyEinoTransientRetrySegment(conversationID, result, curHistory, curFinalMessage, segmentUserMessage)
-	if sendProgress != nil {
-		sendProgress("已恢复上下文，正在继续推理…", map[string]interface{}{
-			"conversationId": conversationID,
-			"source":         "empty_response_continue",
-		})
-	}
-	return true, false
-}

internal/handler/eino_single_agent.go

@@ -119,7 +119,6 @@ func (h *AgentHandler) EinoSingleAgentLoopStream(c *gin.Context) {
 
 	var cancelWithCause context.CancelCauseFunc
 	curFinalMessage := prep.FinalMessage
-	segmentUserMessage := prep.FinalMessage // 本请求原始用户句，临时重试时不得丢失
 	curHistory := prep.History
 	roleTools := prep.RoleTools
 
@@ -177,8 +176,6 @@ func (h *AgentHandler) EinoSingleAgentLoopStream(c *gin.Context) {
 	taskOwned = true
 
 	var cumulativeMCPExecutionIDs []string
-	var transientRunAttempts int
-	var emptyResponseAttempts int
 	// 同一请求内分段续跑时，主代理 iteration 事件按偏移累计，避免 UI 出现「第3轮 → 第1轮」回跳。
 	var mainIterationOffset int
 
@@ -215,6 +212,7 @@ func (h *AgentHandler) EinoSingleAgentLoopStream(c *gin.Context) {
 		}
 		taskCtxLoop := mcp.WithMCPConversationID(taskCtx, conversationID)
 		taskCtxLoop = mcp.WithToolRunRegistry(taskCtxLoop, h.tasks)
+		taskCtxLoop = mcp.WithEinoExecuteRunRegistry(taskCtxLoop, h.tasks)
 		taskCtxLoop = multiagent.WithHITLToolInterceptor(taskCtxLoop, func(ctx context.Context, toolName, arguments string) (string, error) {
 			return h.interceptHITLForEinoTool(ctx, cancelWithCause, conversationID, assistantMessageID, sendEvent, toolName, arguments)
 		})
@@ -233,61 +231,18 @@ func (h *AgentHandler) EinoSingleAgentLoopStream(c *gin.Context) {
 			roleTools,
 			progressCallback,
 			chatReasoningToClientIntent(req.Reasoning),
-			h.projectBlackboardBlock(conversationID),
+			h.agentSessionContextBlock(conversationID),
 		)
 
 		if result != nil && len(result.MCPExecutionIDs) > 0 {
 			cumulativeMCPExecutionIDs = mergeMCPExecutionIDLists(cumulativeMCPExecutionIDs, result.MCPExecutionIDs)
 		}
 
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
-			baseCtx, conversationID, result, runErr, &emptyResponseAttempts,
-			&curHistory, &curFinalMessage, segmentUserMessage, progressCallback,
-			func(msg string, extra map[string]interface{}) { sendEvent("progress", msg, extra) },
-		)
-		if exhaustedEmpty {
-			runErr = nil
-			transientRunAttempts = 0
-			timeoutCancel()
-			break
-		}
-		if handledEmpty {
-			mainIterationOffset += segmentMainIterationMax
-			transientRunAttempts = 0
-			timeoutCancel()
-			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
-			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
-			taskCtx, timeoutCancel = context.WithTimeout(baseCtx, 600*time.Minute)
-			h.tasks.UpdateTaskStatus(conversationID, "running")
-			continue
-		}
-
 		if runErr == nil {
-			// 任一段成功完成后，重置临时错误重试窗口（次数/退避从头开始）。
-			transientRunAttempts = 0
-			emptyResponseAttempts = 0
 			timeoutCancel()
 			break
 		}
 
-		handled, fatalErr := h.handleEinoTransientRetryContinue(
-			baseCtx, conversationID, result, runErr, &transientRunAttempts,
-			&curHistory, &curFinalMessage, segmentUserMessage, progressCallback,
-			func(msg string, extra map[string]interface{}) { sendEvent("progress", msg, extra) },
-		)
-		if handled {
-			mainIterationOffset += segmentMainIterationMax
-			timeoutCancel()
-			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
-			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
-			taskCtx, timeoutCancel = context.WithTimeout(baseCtx, 600*time.Minute)
-			h.tasks.UpdateTaskStatus(conversationID, "running")
-			continue
-		}
-		if fatalErr != nil {
-			runErr = fatalErr
-		}
-
 		cause := context.Cause(baseCtx)
 		if errors.Is(cause, multiagent.ErrInterruptContinue) {
 			if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
@@ -312,8 +267,6 @@ func (h *AgentHandler) EinoSingleAgentLoopStream(c *gin.Context) {
 				"source":         "interrupt_continue",
 			})
 			mainIterationOffset += segmentMainIterationMax
-			// 非临时错误分段续跑（用户中断并继续）时，清空 transient 计数，避免跨分段累加。
-			transientRunAttempts = 0
 			timeoutCancel()
 			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
 			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
@@ -448,8 +401,6 @@ func (h *AgentHandler) EinoSingleAgentLoop(c *gin.Context) {
 	curMsg := prep.FinalMessage
 	var result *multiagent.RunResult
 	var runErr error
-	var transientRunAttempts int
-	var emptyResponseAttempts int
 	for {
 		result, runErr = multiagent.RunEinoSingleChatModelAgent(
 			taskCtx,
@@ -465,30 +416,11 @@ func (h *AgentHandler) EinoSingleAgentLoop(c *gin.Context) {
 			prep.RoleTools,
 			progressCallback,
 			chatReasoningToClientIntent(req.Reasoning),
-			h.projectBlackboardBlock(prep.ConversationID),
+			h.agentSessionContextBlock(prep.ConversationID),
 		)
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
-			baseCtx, prep.ConversationID, result, runErr, &emptyResponseAttempts,
-			&curHist, &curMsg, prep.FinalMessage, progressCallback, nil,
-		)
-		if exhaustedEmpty {
-			runErr = nil
-			break
-		}
-		if handledEmpty {
-			continue
-		}
 		if runErr == nil {
 			break
 		}
-		if handled, fatalErr := h.handleEinoTransientRetryContinue(
-			baseCtx, prep.ConversationID, result, runErr, &transientRunAttempts,
-			&curHist, &curMsg, prep.FinalMessage, progressCallback, nil,
-		); handled {
-			continue
-		} else if fatalErr != nil {
-			runErr = fatalErr
-		}
 		if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
 			h.persistEinoAgentTraceForResume(prep.ConversationID, result)
 		}

internal/handler/monitor.go

@@ -10,8 +10,10 @@ import (
 	"time"
 
 	"cyberstrike-ai/internal/audit"
+	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/database"
 	"cyberstrike-ai/internal/mcp"
+	"cyberstrike-ai/internal/monitor"
 	"cyberstrike-ai/internal/security"
 	"github.com/gin-gonic/gin"
 	"go.uber.org/zap"
@@ -21,10 +23,18 @@ import (
 type MonitorHandler struct {
 	mcpServer        *mcp.Server
 	externalMCPMgr   *mcp.ExternalMCPManager
+	taskManager      *AgentTaskManager
+	agentHandler     *AgentHandler
 	executor         *security.Executor
 	db               *database.DB
 	logger           *zap.Logger
 	audit            *audit.Service
+	monitorRetention *monitor.Service
+}
+
+// SetMonitorRetention wires MCP execution retention settings.
+func (h *MonitorHandler) SetMonitorRetention(s *monitor.Service) {
+	h.monitorRetention = s
 }
 
 // SetAudit wires platform audit logging.
@@ -48,6 +58,16 @@ func (h *MonitorHandler) SetExternalMCPManager(mgr *mcp.ExternalMCPManager) {
 	h.externalMCPMgr = mgr
 }
 
+// SetTaskManager 设置 Agent 任务管理器（用于 Eino execute 等按 executionId 终止）。
+func (h *MonitorHandler) SetTaskManager(mgr *AgentTaskManager) {
+	h.taskManager = mgr
+}
+
+// SetAgentHandler 设置 Agent 处理器（MCP 监控终止与对话页「中断并继续」共用逻辑）。
+func (h *MonitorHandler) SetAgentHandler(ah *AgentHandler) {
+	h.agentHandler = ah
+}
+
 // MonitorResponse 监控响应
 type MonitorResponse struct {
 	Executions    []*mcp.ToolExecution      `json:"executions"`
@@ -57,6 +77,7 @@ type MonitorResponse struct {
 	Page          int                       `json:"page,omitempty"`
 	PageSize      int                       `json:"page_size,omitempty"`
 	TotalPages    int                       `json:"total_pages,omitempty"`
+	RetentionDays int                       `json:"retention_days,omitempty"`
 }
 
 // Monitor 获取监控信息
@@ -81,6 +102,7 @@ func (h *MonitorHandler) Monitor(c *gin.Context) {
 	toolName := normalizeToolNameFilter(c.Query("tool"))
 
 	executions, total := h.loadExecutionsWithPagination(page, pageSize, status, toolName)
+	h.enrichExecutionsConversationID(executions)
 	stats := h.loadStats()
 
 	totalPages := (total + pageSize - 1) / pageSize
@@ -96,9 +118,17 @@ func (h *MonitorHandler) Monitor(c *gin.Context) {
 		Page:          page,
 		PageSize:      pageSize,
 		TotalPages:    totalPages,
+		RetentionDays: h.monitorRetentionDays(),
 	})
 }
 
+func (h *MonitorHandler) monitorRetentionDays() int {
+	if h.monitorRetention != nil {
+		return h.monitorRetention.RetentionDays()
+	}
+	return config.MonitorConfig{}.RetentionDaysEffective()
+}
+
 func (h *MonitorHandler) loadExecutions() []*mcp.ToolExecution {
 	executions, _ := h.loadExecutionsWithPagination(1, 1000, "", "")
 	return executions
@@ -230,6 +260,7 @@ func (h *MonitorHandler) GetExecution(c *gin.Context) {
 	// 先从内部MCP服务器查找
 	exec, exists := h.mcpServer.GetExecution(id)
 	if exists {
+		h.enrichExecutionsConversationID([]*mcp.ToolExecution{exec})
 		c.JSON(http.StatusOK, exec)
 		return
 	}
@@ -238,6 +269,7 @@ func (h *MonitorHandler) GetExecution(c *gin.Context) {
 	if h.externalMCPMgr != nil {
 		exec, exists = h.externalMCPMgr.GetExecution(id)
 		if exists {
+			h.enrichExecutionsConversationID([]*mcp.ToolExecution{exec})
 			c.JSON(http.StatusOK, exec)
 			return
 		}
@@ -247,6 +279,7 @@ func (h *MonitorHandler) GetExecution(c *gin.Context) {
 	if h.db != nil {
 		exec, err := h.db.GetToolExecution(id)
 		if err == nil && exec != nil {
+			h.enrichExecutionsConversationID([]*mcp.ToolExecution{exec})
 			c.JSON(http.StatusOK, exec)
 			return
 		}
@@ -273,6 +306,19 @@ func (h *MonitorHandler) CancelExecution(c *gin.Context) {
 		return
 	}
 	note = strings.TrimSpace(body.Note)
+
+	convID := h.conversationIDForRunningExecution(id)
+	if convID != "" && h.agentHandler != nil {
+		if ok, payload := h.agentHandler.cancelToolContinueAfter(convID, id, note); ok {
+			h.logger.Info("MCP 监控页终止工具（与对话中断并继续一致）",
+				zap.String("executionId", id),
+				zap.String("conversationId", convID),
+				zap.Bool("hasNote", note != ""),
+			)
+			c.JSON(http.StatusOK, payload)
+			return
+		}
+	}
 	if h.mcpServer.CancelToolExecutionWithNote(id, note) {
 		h.logger.Info("已请求取消 MCP 工具执行", zap.String("executionId", id), zap.String("source", "internal"), zap.Bool("hasNote", note != ""))
 		c.JSON(http.StatusOK, gin.H{"message": "已发送终止信号", "executionId": id})
@@ -286,6 +332,52 @@ func (h *MonitorHandler) CancelExecution(c *gin.Context) {
 	c.JSON(http.StatusNotFound, gin.H{"error": "未找到进行中的工具执行，或该任务已结束"})
 }
 
+func (h *MonitorHandler) enrichExecutionsConversationID(executions []*mcp.ToolExecution) {
+	for _, exec := range executions {
+		if exec == nil {
+			continue
+		}
+		exec.ConversationID = h.conversationIDForRunningExecution(exec.ID)
+	}
+}
+
+func (h *MonitorHandler) conversationIDForRunningExecution(executionID string) string {
+	executionID = strings.TrimSpace(executionID)
+	if executionID == "" || h.taskManager == nil {
+		return ""
+	}
+	if conv := h.taskManager.ConversationIDForActiveMCPExecution(executionID); conv != "" {
+		return conv
+	}
+	exec := h.lookupExecution(executionID)
+	if exec == nil || exec.Status != "running" {
+		return ""
+	}
+	if strings.TrimSpace(exec.ToolName) == "execute" {
+		if onlyConv, ok := h.taskManager.ConversationIDForActiveEinoExecute(); ok {
+			return onlyConv
+		}
+	}
+	return ""
+}
+
+func (h *MonitorHandler) lookupExecution(id string) *mcp.ToolExecution {
+	if exec, ok := h.mcpServer.GetExecution(id); ok {
+		return exec
+	}
+	if h.externalMCPMgr != nil {
+		if exec, ok := h.externalMCPMgr.GetExecution(id); ok {
+			return exec
+		}
+	}
+	if h.db != nil {
+		if exec, err := h.db.GetToolExecution(id); err == nil && exec != nil {
+			return exec
+		}
+	}
+	return nil
+}
+
 // BatchGetToolNames 批量获取工具执行的工具名称（消除前端 N+1 请求）
 func (h *MonitorHandler) BatchGetToolNames(c *gin.Context) {
 	var req struct {

internal/handler/multi_agent.go

@@ -136,7 +136,6 @@ func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
 
 	var cancelWithCause context.CancelCauseFunc
 	curFinalMessage := prep.FinalMessage
-	segmentUserMessage := prep.FinalMessage // 本请求原始用户句，临时重试时不得丢失
 	curHistory := prep.History
 	roleTools := prep.RoleTools
 	orch := strings.TrimSpace(req.Orchestration)
@@ -187,8 +186,6 @@ func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
 
 	// 同一 HTTP 流内多段 Run（如中断并继续）合并 MCP execution id，供最终 response / 库表与工具芯片展示完整列表
 	var cumulativeMCPExecutionIDs []string
-	var transientRunAttempts int
-	var emptyResponseAttempts int
 	// 同一请求内分段续跑时，主代理 iteration 事件按偏移累计，避免 UI 出现「第3轮 → 第1轮」回跳。
 	var mainIterationOffset int
 
@@ -225,6 +222,7 @@ func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
 		}
 		taskCtxLoop := mcp.WithMCPConversationID(taskCtx, conversationID)
 		taskCtxLoop = mcp.WithToolRunRegistry(taskCtxLoop, h.tasks)
+		taskCtxLoop = mcp.WithEinoExecuteRunRegistry(taskCtxLoop, h.tasks)
 		taskCtxLoop = multiagent.WithHITLToolInterceptor(taskCtxLoop, func(ctx context.Context, toolName, arguments string) (string, error) {
 			return h.interceptHITLForEinoTool(ctx, cancelWithCause, conversationID, assistantMessageID, sendEvent, toolName, arguments)
 		})
@@ -245,61 +243,18 @@ func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
 			h.agentsMarkdownDir,
 			orch,
 			chatReasoningToClientIntent(req.Reasoning),
-			h.projectBlackboardBlock(conversationID),
+			h.agentSessionContextBlock(conversationID),
 		)
 
 		if result != nil && len(result.MCPExecutionIDs) > 0 {
 			cumulativeMCPExecutionIDs = mergeMCPExecutionIDLists(cumulativeMCPExecutionIDs, result.MCPExecutionIDs)
 		}
 
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
-			baseCtx, conversationID, result, runErr, &emptyResponseAttempts,
-			&curHistory, &curFinalMessage, segmentUserMessage, progressCallback,
-			func(msg string, extra map[string]interface{}) { sendEvent("progress", msg, extra) },
-		)
-		if exhaustedEmpty {
-			runErr = nil
-			transientRunAttempts = 0
-			timeoutCancel()
-			break
-		}
-		if handledEmpty {
-			mainIterationOffset += segmentMainIterationMax
-			transientRunAttempts = 0
-			timeoutCancel()
-			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
-			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
-			taskCtx, timeoutCancel = context.WithTimeout(baseCtx, 600*time.Minute)
-			h.tasks.UpdateTaskStatus(conversationID, "running")
-			continue
-		}
-
 		if runErr == nil {
-			// 任一段成功完成后，重置临时错误重试窗口（次数/退避从头开始）。
-			transientRunAttempts = 0
-			emptyResponseAttempts = 0
 			timeoutCancel()
 			break
 		}
 
-		handled, fatalErr := h.handleEinoTransientRetryContinue(
-			baseCtx, conversationID, result, runErr, &transientRunAttempts,
-			&curHistory, &curFinalMessage, segmentUserMessage, progressCallback,
-			func(msg string, extra map[string]interface{}) { sendEvent("progress", msg, extra) },
-		)
-		if handled {
-			mainIterationOffset += segmentMainIterationMax
-			timeoutCancel()
-			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
-			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
-			taskCtx, timeoutCancel = context.WithTimeout(baseCtx, 600*time.Minute)
-			h.tasks.UpdateTaskStatus(conversationID, "running")
-			continue
-		}
-		if fatalErr != nil {
-			runErr = fatalErr
-		}
-
 		cause := context.Cause(baseCtx)
 		if errors.Is(cause, multiagent.ErrInterruptContinue) {
 			if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
@@ -324,8 +279,6 @@ func (h *AgentHandler) MultiAgentLoopStream(c *gin.Context) {
 				"source":         "interrupt_continue",
 			})
 			mainIterationOffset += segmentMainIterationMax
-			// 非临时错误分段续跑（用户中断并继续）时，清空 transient 计数，避免跨分段累加。
-			transientRunAttempts = 0
 			timeoutCancel()
 			baseCtx, cancelWithCause = context.WithCancelCause(context.Background())
 			h.tasks.BindTaskCancel(conversationID, cancelWithCause)
@@ -460,8 +413,6 @@ func (h *AgentHandler) MultiAgentLoop(c *gin.Context) {
 	curMsg := prep.FinalMessage
 	var result *multiagent.RunResult
 	var runErr error
-	var transientRunAttempts int
-	var emptyResponseAttempts int
 	for {
 		result, runErr = multiagent.RunDeepAgent(
 			taskCtx,
@@ -479,30 +430,11 @@ func (h *AgentHandler) MultiAgentLoop(c *gin.Context) {
 			h.agentsMarkdownDir,
 			strings.TrimSpace(req.Orchestration),
 			chatReasoningToClientIntent(req.Reasoning),
-			h.projectBlackboardBlock(prep.ConversationID),
+			h.agentSessionContextBlock(prep.ConversationID),
 		)
-		handledEmpty, exhaustedEmpty := h.handleEinoEmptyResponseContinue(
-			baseCtx, prep.ConversationID, result, runErr, &emptyResponseAttempts,
-			&curHist, &curMsg, prep.FinalMessage, progressCallback, nil,
-		)
-		if exhaustedEmpty {
-			runErr = nil
-			break
-		}
-		if handledEmpty {
-			continue
-		}
 		if runErr == nil {
 			break
 		}
-		if handled, fatalErr := h.handleEinoTransientRetryContinue(
-			baseCtx, prep.ConversationID, result, runErr, &transientRunAttempts,
-			&curHist, &curMsg, prep.FinalMessage, progressCallback, nil,
-		); handled {
-			continue
-		} else if fatalErr != nil {
-			runErr = fatalErr
-		}
 		if shouldPersistEinoAgentTraceAfterRunError(baseCtx) {
 			h.persistEinoAgentTraceForResume(prep.ConversationID, result)
 		}

internal/handler/project_context.go

@@ -7,6 +7,42 @@ import (
 	"go.uber.org/zap"
 )
 
+// agentSessionContextBlock 注入会话工作目录与项目黑板（用于 system prompt 追加块）。
+func (h *AgentHandler) agentSessionContextBlock(conversationID string) string {
+	var parts []string
+	if ws := h.buildWorkspaceBlock(conversationID); ws != "" {
+		parts = append(parts, ws)
+	}
+	if bb := h.projectBlackboardBlock(conversationID); bb != "" {
+		parts = append(parts, bb)
+	}
+	return strings.Join(parts, "\n\n")
+}
+
+func (h *AgentHandler) buildWorkspaceBlock(conversationID string) string {
+	if h == nil || h.config == nil {
+		return ""
+	}
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" {
+		return ""
+	}
+	projectID := h.conversationProjectID(conversationID)
+	rel := project.WorkspaceRootDir(h.config.Agent.WorkspaceRootDir, projectID, conversationID)
+	abs, err := project.EnsureWorkspace(rel)
+	if err != nil {
+		if h.logger != nil {
+			h.logger.Warn("创建会话工作目录失败",
+				zap.String("conversationId", conversationID),
+				zap.String("projectId", projectID),
+				zap.String("path", rel),
+				zap.Error(err))
+		}
+		return ""
+	}
+	return project.BuildWorkspaceBlock(abs)
+}
+
 // projectBlackboardBlock 根据对话 ID 构建项目事实索引块（用于注入 system prompt）。
 func (h *AgentHandler) projectBlackboardBlock(conversationID string) string {
 	if h == nil || h.db == nil || h.config == nil {

internal/handler/robot.go

@@ -594,6 +594,9 @@ func (h *RobotHandler) cmdDelete(platform, userID, convID string) string {
 		h.mu.Unlock()
 		h.deleteSessionBinding(sk)
 	}
+	if h.agentHandler != nil {
+		h.agentHandler.CancelRunningTaskForConversation(convID)
+	}
 	if err := h.db.DeleteConversation(convID); err != nil {
 		return "删除失败: " + err.Error()
 	}

internal/handler/task_manager.go

@@ -37,6 +37,11 @@ type AgentTask struct {
 	// InterruptContinueNote 无 MCP 时「中断并继续」由用户在弹窗中填写的补充说明（Cancel 前写入，续跑轮次读取后清空）
 	InterruptContinueNote string `json:"-"`
 
+	// activeEinoExecuteCancel 当前进行中的 Eino filesystem execute 取消函数（与 MCP 工具并行，供中断并继续）
+	activeEinoExecuteCancel context.CancelFunc
+	// activeEinoExecuteAbortNote AbortActiveEinoExecute 写入的用户说明，由 execute 收尾时合并进工具结果
+	activeEinoExecuteAbortNote string
+
 	cancel func(error)
 }
 
@@ -70,6 +75,103 @@ func (m *AgentTaskManager) UnregisterRunningTool(conversationID, executionID str
 	}
 }
 
+// RegisterActiveEinoExecute 登记进行中的 Eino filesystem execute（每会话同时仅一条）。
+func (m *AgentTaskManager) RegisterActiveEinoExecute(conversationID string, cancel context.CancelFunc) {
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" || cancel == nil {
+		return
+	}
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	if t, ok := m.tasks[conversationID]; ok && t != nil {
+		t.activeEinoExecuteCancel = cancel
+		t.activeEinoExecuteAbortNote = ""
+	}
+}
+
+// UnregisterActiveEinoExecute execute 正常结束或已取消后清除登记。
+func (m *AgentTaskManager) UnregisterActiveEinoExecute(conversationID string) {
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" {
+		return
+	}
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	if t, ok := m.tasks[conversationID]; ok && t != nil {
+		t.activeEinoExecuteCancel = nil
+		t.activeEinoExecuteAbortNote = ""
+	}
+}
+
+// ConversationIDForActiveMCPExecution 根据当前登记的工具 executionId 反查会话 ID（供 MCP 监控页按 executionId 终止）。
+func (m *AgentTaskManager) ConversationIDForActiveMCPExecution(executionID string) string {
+	executionID = strings.TrimSpace(executionID)
+	if executionID == "" {
+		return ""
+	}
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	for convID, t := range m.tasks {
+		if t != nil && t.ActiveMCPExecutionID == executionID {
+			return convID
+		}
+	}
+	return ""
+}
+
+// ConversationIDForActiveEinoExecute 返回当前唯一进行 Eino execute 的会话 ID；多会话并行时返回空。
+func (m *AgentTaskManager) ConversationIDForActiveEinoExecute() (string, bool) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	var found string
+	count := 0
+	for convID, t := range m.tasks {
+		if t != nil && t.activeEinoExecuteCancel != nil {
+			found = convID
+			count++
+		}
+	}
+	if count == 1 {
+		return found, true
+	}
+	return "", false
+}
+
+// AbortActiveEinoExecute 终止当前 Eino execute 并暂存用户说明（与 MCP 工具终止一致）。
+func (m *AgentTaskManager) AbortActiveEinoExecute(conversationID, note string) bool {
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" {
+		return false
+	}
+	m.mu.Lock()
+	t, ok := m.tasks[conversationID]
+	if !ok || t == nil || t.activeEinoExecuteCancel == nil {
+		m.mu.Unlock()
+		return false
+	}
+	t.activeEinoExecuteAbortNote = strings.TrimSpace(note)
+	cancel := t.activeEinoExecuteCancel
+	m.mu.Unlock()
+	cancel()
+	return true
+}
+
+// TakeEinoExecuteAbortNote 读取并清空 execute 终止说明（execute 收尾时调用一次）。
+func (m *AgentTaskManager) TakeEinoExecuteAbortNote(conversationID string) string {
+	conversationID = strings.TrimSpace(conversationID)
+	if conversationID == "" {
+		return ""
+	}
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	if t, ok := m.tasks[conversationID]; ok && t != nil {
+		n := t.activeEinoExecuteAbortNote
+		t.activeEinoExecuteAbortNote = ""
+		return n
+	}
+	return ""
+}
+
 // SetInterruptContinueNote 在发起 ErrInterruptContinue 取消前写入用户补充说明（仅内存）。
 func (m *AgentTaskManager) SetInterruptContinueNote(conversationID, note string) {
 	conversationID = strings.TrimSpace(conversationID)

internal/handler/task_manager_eino_execute_test.go

@@ -0,0 +1,56 @@
+package handler
+
+import (
+	"context"
+	"testing"
+	"time"
+)
+
+func TestAbortActiveEinoExecute(t *testing.T) {
+	m := NewAgentTaskManager()
+	conv := "conv-eino-exec-abort"
+	ctx, cancel := context.WithCancel(context.Background())
+	_, err := m.StartTask(conv, "test", func(error) {})
+	if err != nil {
+		t.Fatalf("StartTask: %v", err)
+	}
+	m.RegisterActiveEinoExecute(conv, cancel)
+
+	done := make(chan struct{})
+	go func() {
+		<-ctx.Done()
+		close(done)
+	}()
+
+	if !m.AbortActiveEinoExecute(conv, "跳过域名收集") {
+		t.Fatal("expected abort to succeed")
+	}
+	select {
+	case <-done:
+	case <-time.After(2 * time.Second):
+		t.Fatal("execute cancel did not propagate")
+	}
+	if got := m.TakeEinoExecuteAbortNote(conv); got != "跳过域名收集" {
+		t.Fatalf("abort note = %q, want 跳过域名收集", got)
+	}
+	m.UnregisterActiveEinoExecute(conv)
+	if m.AbortActiveEinoExecute(conv, "") {
+		t.Fatal("second abort should fail when no active execute")
+	}
+}
+
+func TestConversationIDForActiveMCPExecution(t *testing.T) {
+	m := NewAgentTaskManager()
+	conv := "conv-mcp-exec"
+	_, err := m.StartTask(conv, "test", func(error) {})
+	if err != nil {
+		t.Fatalf("StartTask: %v", err)
+	}
+	m.RegisterRunningTool(conv, "exec-123")
+	if got := m.ConversationIDForActiveMCPExecution("exec-123"); got != conv {
+		t.Fatalf("got %q, want %q", got, conv)
+	}
+	if got := m.ConversationIDForActiveMCPExecution("missing"); got != "" {
+		t.Fatalf("missing should be empty, got %q", got)
+	}
+}

internal/mcp/run_context.go

@@ -11,7 +11,16 @@ type ToolRunRegistry interface {
 	UnregisterRunningTool(conversationID, executionID string)
 }
 
+// EinoExecuteRunRegistry 登记进行中的 Eino filesystem execute，供「中断并继续」终止 amass 等长命令。
+type EinoExecuteRunRegistry interface {
+	RegisterActiveEinoExecute(conversationID string, cancel context.CancelFunc)
+	UnregisterActiveEinoExecute(conversationID string)
+	AbortActiveEinoExecute(conversationID, note string) bool
+	TakeEinoExecuteAbortNote(conversationID string) string
+}
+
 type toolRunRegistryCtxKey struct{}
+type einoExecuteRunRegistryCtxKey struct{}
 type mcpConversationIDCtxKey struct{}
 
 // WithToolRunRegistry 将登记器注入 ctx（Eino / 原生 Agent 任务 ctx）。
@@ -31,6 +40,23 @@ func ToolRunRegistryFromContext(ctx context.Context) ToolRunRegistry {
 	return v
 }
 
+// WithEinoExecuteRunRegistry 将 Eino execute 取消登记器注入 ctx。
+func WithEinoExecuteRunRegistry(ctx context.Context, reg EinoExecuteRunRegistry) context.Context {
+	if ctx == nil || reg == nil {
+		return ctx
+	}
+	return context.WithValue(ctx, einoExecuteRunRegistryCtxKey{}, reg)
+}
+
+// EinoExecuteRunRegistryFromContext 取出 Eino execute 登记器（无则 nil）。
+func EinoExecuteRunRegistryFromContext(ctx context.Context) EinoExecuteRunRegistry {
+	if ctx == nil {
+		return nil
+	}
+	v, _ := ctx.Value(einoExecuteRunRegistryCtxKey{}).(EinoExecuteRunRegistry)
+	return v
+}
+
 // WithMCPConversationID 将对话 ID 注入 ctx，供 CallTool 内与 executionId 关联。
 func WithMCPConversationID(ctx context.Context, conversationID string) context.Context {
 	if ctx == nil {

internal/mcp/server.go

@@ -921,9 +921,8 @@ func (s *Server) CallTool(ctx context.Context, toolName string, args map[string]
 	return finalResult, executionID, nil
 }
 
-// RecordCompletedToolInvocation 将已在其它路径完成的工具调用写入监控存储（格式与 CallTool 结束后一致），
+// BeginToolExecution 创建 running 状态的执行记录，供 Eino 等非 CallTool 路径在工具开始时落库。
-// 用于 Eino ADK filesystem execute 等未经过 CallTool 的场景；返回 executionId 供助手消息 mcpExecutionIds 关联。
+func (s *Server) BeginToolExecution(toolName string, args map[string]interface{}) string {
-func (s *Server) RecordCompletedToolInvocation(toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
 	if s == nil {
 		return ""
 	}
@@ -931,21 +930,73 @@ func (s *Server) RecordCompletedToolInvocation(toolName string, args map[string]
 		args = map[string]interface{}{}
 	}
 	executionID := uuid.New().String()
-	now := time.Now()
+	execution := &ToolExecution{
-	failed := invokeErr != nil
-	exec := &ToolExecution{
 		ID:        executionID,
 		ToolName:  toolName,
 		Arguments: args,
-		StartTime: now,
+		Status:    "running",
-		EndTime:   &now,
+		StartTime: time.Now(),
-		Duration:  0,
 	}
+
+	s.mu.Lock()
+	s.executions[executionID] = execution
+	s.cleanupOldExecutions()
+	s.mu.Unlock()
+
+	if s.storage != nil {
+		if err := s.storage.SaveToolExecution(execution); err != nil {
+			s.logger.Warn("保存执行记录到数据库失败", zap.Error(err))
+		}
+	}
+	return executionID
+}
+
+// FinishToolExecution 完成先前 BeginToolExecution 创建的记录；executionID 为空时等同 RecordCompletedToolInvocation。
+func (s *Server) FinishToolExecution(executionID, toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
+	if s == nil {
+		return ""
+	}
+	if args == nil {
+		args = map[string]interface{}{}
+	}
+	id := strings.TrimSpace(executionID)
+	if id == "" {
+		return s.RecordCompletedToolInvocation(toolName, args, resultText, invokeErr)
+	}
+
+	now := time.Now()
+	failed := invokeErr != nil
+	var finalResult *ToolResult
+
+	s.mu.Lock()
+	exec, inMem := s.executions[id]
+	if !inMem || exec == nil {
+		exec = &ToolExecution{
+			ID:        id,
+			ToolName:  toolName,
+			Arguments: args,
+			StartTime: now,
+		}
+		s.executions[id] = exec
+	} else if toolName != "" {
+		exec.ToolName = toolName
+	}
+	if len(args) > 0 {
+		exec.Arguments = args
+	}
+	exec.EndTime = &now
+	if exec.StartTime.IsZero() {
+		exec.StartTime = now
+	}
+	exec.Duration = now.Sub(exec.StartTime)
+
 	if failed {
-		exec.Status = "failed"
+		st, msg := executionStatusAndMessage(invokeErr)
-		exec.Error = invokeErr.Error()
+		exec.Status = st
+		exec.Error = msg
 		if strings.TrimSpace(resultText) != "" {
-			exec.Result = &ToolResult{Content: []Content{{Type: "text", Text: resultText}}}
+			finalResult = &ToolResult{Content: []Content{{Type: "text", Text: resultText}}}
+			exec.Result = finalResult
 		}
 	} else {
 		exec.Status = "completed"
@@ -953,15 +1004,31 @@ func (s *Server) RecordCompletedToolInvocation(toolName string, args map[string]
 		if strings.TrimSpace(text) == "" {
 			text = "（无输出）"
 		}
-		exec.Result = &ToolResult{Content: []Content{{Type: "text", Text: text}}}
+		finalResult = &ToolResult{Content: []Content{{Type: "text", Text: text}}}
+		exec.Result = finalResult
 	}
+	s.mu.Unlock()
+
 	if s.storage != nil {
 		if err := s.storage.SaveToolExecution(exec); err != nil {
-			s.logger.Warn("RecordCompletedToolInvocation 保存失败", zap.Error(err))
+			s.logger.Warn("保存执行记录到数据库失败", zap.Error(err))
 		}
 	}
-	s.updateStats(toolName, failed)
+
-	return executionID
+	s.updateStats(exec.ToolName, failed)
+
+	if s.storage != nil {
+		s.mu.Lock()
+		delete(s.executions, id)
+		s.mu.Unlock()
+	}
+	return id
+}
+
+// RecordCompletedToolInvocation 将已在其它路径完成的工具调用写入监控存储（格式与 CallTool 结束后一致），
+// 用于 Eino ADK filesystem execute 等未经过 CallTool 的场景；返回 executionId 供助手消息 mcpExecutionIds 关联。
+func (s *Server) RecordCompletedToolInvocation(toolName string, args map[string]interface{}, resultText string, invokeErr error) string {
+	return s.FinishToolExecution("", toolName, args, resultText, invokeErr)
 }
 
 // UpdateToolExecutionResult 将监控库中的工具结果更新为送入模型的展示正文（如 reduction 后的 persisted-output）。

internal/mcp/types.go

@@ -199,6 +199,8 @@ type ToolExecution struct {
 	StartTime time.Time              `json:"startTime"`
 	EndTime   *time.Time             `json:"endTime,omitempty"`
 	Duration  time.Duration          `json:"duration,omitempty"`
+	// ConversationID 仅 API 展示用（进行中的 Agent 任务），不写入 tool_executions 表。
+	ConversationID string `json:"conversationId,omitempty"`
 }
 
 // ToolStats 工具统计信息

internal/monitor/retention.go

@@ -0,0 +1,71 @@
+package monitor
+
+import (
+	"time"
+
+	"cyberstrike-ai/internal/config"
+	"cyberstrike-ai/internal/database"
+
+	"go.uber.org/zap"
+)
+
+const retentionPurgeInterval = time.Hour
+
+// Service manages MCP tool execution monitor retention.
+type Service struct {
+	db     *database.DB
+	cfg    *config.Config
+	logger *zap.Logger
+}
+
+// NewService creates a monitor retention service.
+func NewService(db *database.DB, cfg *config.Config, logger *zap.Logger) *Service {
+	return &Service{db: db, cfg: cfg, logger: logger}
+}
+
+// RetentionDays returns configured retention; 0 means keep forever.
+func (s *Service) RetentionDays() int {
+	if s == nil || s.cfg == nil {
+		return config.MonitorConfig{}.RetentionDaysEffective()
+	}
+	return s.cfg.Monitor.RetentionDaysEffective()
+}
+
+// PurgeExpired deletes tool execution rows older than retention_days when configured.
+func (s *Service) PurgeExpired() {
+	if s == nil || s.db == nil || s.cfg == nil {
+		return
+	}
+	days := s.cfg.Monitor.RetentionDaysEffective()
+	if days <= 0 {
+		return
+	}
+	cutoff := time.Now().AddDate(0, 0, -days)
+	n, err := s.db.PurgeToolExecutionsBefore(cutoff)
+	if err != nil {
+		if s.logger != nil {
+			s.logger.Warn("清理过期 MCP 执行记录失败", zap.Error(err))
+		}
+		return
+	}
+	if n > 0 && s.logger != nil {
+		s.logger.Info("已清理过期 MCP 执行记录", zap.Int64("deleted", n), zap.Int("retention_days", days))
+	}
+}
+
+// StartRetentionLoop periodically purges expired tool execution rows.
+func StartRetentionLoop(s *Service, logger *zap.Logger) {
+	if s == nil {
+		return
+	}
+	go func() {
+		ticker := time.NewTicker(retentionPurgeInterval)
+		defer ticker.Stop()
+		for range ticker.C {
+			s.PurgeExpired()
+			if logger != nil {
+				logger.Debug("monitor retention tick completed")
+			}
+		}
+	}()
+}

internal/monitor/retention_test.go

@@ -0,0 +1,94 @@
+package monitor
+
+import (
+	"path/filepath"
+	"testing"
+	"time"
+
+	"cyberstrike-ai/internal/config"
+	"cyberstrike-ai/internal/database"
+	"cyberstrike-ai/internal/mcp"
+
+	"go.uber.org/zap"
+)
+
+func TestServicePurgeExpired_respectsZeroRetention(t *testing.T) {
+	dbPath := filepath.Join(t.TempDir(), "monitor.db")
+	db, err := database.NewDB(dbPath, zap.NewNop())
+	if err != nil {
+		t.Fatalf("NewDB: %v", err)
+	}
+	defer db.Close()
+
+	exec := &mcp.ToolExecution{
+		ID:        "ancient",
+		ToolName:  "curl::get",
+		Arguments: map[string]interface{}{},
+		Status:    "completed",
+		StartTime: mustParseTime(t, "2020-01-01T00:00:00Z"),
+	}
+	if err := db.SaveToolExecution(exec); err != nil {
+		t.Fatalf("SaveToolExecution: %v", err)
+	}
+
+	zero := 0
+	svc := NewService(db, &config.Config{
+		Monitor: config.MonitorConfig{RetentionDays: &zero},
+	}, zap.NewNop())
+	svc.PurgeExpired()
+
+	if _, err := db.GetToolExecution("ancient"); err != nil {
+		t.Fatalf("record should remain when retention_days=0: %v", err)
+	}
+}
+
+func TestServicePurgeExpired_deletesOldRows(t *testing.T) {
+	dbPath := filepath.Join(t.TempDir(), "monitor.db")
+	db, err := database.NewDB(dbPath, zap.NewNop())
+	if err != nil {
+		t.Fatalf("NewDB: %v", err)
+	}
+	defer db.Close()
+
+	exec := &mcp.ToolExecution{
+		ID:        "ancient",
+		ToolName:  "curl::get",
+		Arguments: map[string]interface{}{},
+		Status:    "completed",
+		StartTime: mustParseTime(t, "2020-01-01T00:00:00Z"),
+	}
+	if err := db.SaveToolExecution(exec); err != nil {
+		t.Fatalf("SaveToolExecution: %v", err)
+	}
+
+	days := 90
+	svc := NewService(db, &config.Config{
+		Monitor: config.MonitorConfig{RetentionDays: &days},
+	}, zap.NewNop())
+	svc.PurgeExpired()
+
+	if _, err := db.GetToolExecution("ancient"); err == nil {
+		t.Fatal("record should be purged when older than retention_days")
+	}
+}
+
+func TestRetentionDaysEffective_defaults(t *testing.T) {
+	got := config.MonitorConfig{}.RetentionDaysEffective()
+	if got != 90 {
+		t.Fatalf("default = %d, want 90", got)
+	}
+	zero := 0
+	cfg := config.MonitorConfig{RetentionDays: &zero}
+	if cfg.RetentionDaysEffective() != 0 {
+		t.Fatalf("zero = %d, want 0", cfg.RetentionDaysEffective())
+	}
+}
+
+func mustParseTime(t *testing.T, value string) time.Time {
+	t.Helper()
+	parsed, err := time.Parse(time.RFC3339, value)
+	if err != nil {
+		t.Fatalf("parse time: %v", err)
+	}
+	return parsed
+}

internal/multiagent/continuation_user_dedup_middleware.go

@@ -0,0 +1,104 @@
+package multiagent
+
+import (
+	"context"
+	"strings"
+
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/schema"
+	"go.uber.org/zap"
+)
+
+// continuationSessionMarker matches Cursor / IDE session-resume user injections.
+const continuationSessionMarker = "This session is being continued from a previous conversation"
+
+// continuationUserDedupMiddleware keeps only the latest session-resume user message when
+// multiple continuation injections were stacked (e.g. after repeated out-of-context resumes).
+type continuationUserDedupMiddleware struct {
+	adk.BaseChatModelAgentMiddleware
+	logger *zap.Logger
+	phase  string
+}
+
+func newContinuationUserDedupMiddleware(logger *zap.Logger, phase string) adk.ChatModelAgentMiddleware {
+	return &continuationUserDedupMiddleware{logger: logger, phase: phase}
+}
+
+func (m *continuationUserDedupMiddleware) BeforeModelRewriteState(
+	ctx context.Context,
+	state *adk.ChatModelAgentState,
+	mc *adk.ModelContext,
+) (context.Context, *adk.ChatModelAgentState, error) {
+	_ = mc
+	if m == nil || state == nil || len(state.Messages) == 0 {
+		return ctx, state, nil
+	}
+	deduped, dropped := dedupContinuationUserMessages(state.Messages)
+	if dropped == 0 {
+		return ctx, state, nil
+	}
+	if m.logger != nil {
+		m.logger.Info("eino continuation user messages deduplicated",
+			zap.String("phase", m.phase),
+			zap.Int("dropped", dropped),
+			zap.Int("messages_before", len(state.Messages)),
+			zap.Int("messages_after", len(deduped)),
+		)
+	}
+	out := *state
+	out.Messages = deduped
+	return ctx, &out, nil
+}
+
+func adkUserMessageText(msg adk.Message) string {
+	if msg == nil {
+		return ""
+	}
+	var b strings.Builder
+	if s := strings.TrimSpace(msg.Content); s != "" {
+		b.WriteString(s)
+	}
+	for _, part := range msg.UserInputMultiContent {
+		if part.Type == schema.ChatMessagePartTypeText {
+			if s := strings.TrimSpace(part.Text); s != "" {
+				if b.Len() > 0 {
+					b.WriteByte('\n')
+				}
+				b.WriteString(s)
+			}
+		}
+	}
+	return b.String()
+}
+
+func isContinuationUserMessage(msg adk.Message) bool {
+	if msg == nil || msg.Role != schema.User {
+		return false
+	}
+	return strings.Contains(adkUserMessageText(msg), continuationSessionMarker)
+}
+
+func dedupContinuationUserMessages(msgs []adk.Message) ([]adk.Message, int) {
+	lastIdx := -1
+	contCount := 0
+	for i, msg := range msgs {
+		if !isContinuationUserMessage(msg) {
+			continue
+		}
+		contCount++
+		lastIdx = i
+	}
+	if contCount <= 1 {
+		return msgs, 0
+	}
+	out := make([]adk.Message, 0, len(msgs)-(contCount-1))
+	dropped := 0
+	for i, msg := range msgs {
+		if isContinuationUserMessage(msg) && i != lastIdx {
+			dropped++
+			continue
+		}
+		out = append(out, msg)
+	}
+	return out, dropped
+}

internal/multiagent/continuation_user_dedup_middleware_test.go

@@ -0,0 +1,65 @@
+package multiagent
+
+import (
+	"context"
+	"strings"
+	"testing"
+
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/schema"
+)
+
+func continuationUser(text string) adk.Message {
+	return &schema.Message{
+		Role: schema.User,
+		UserInputMultiContent: []schema.MessageInputPart{
+			{Type: schema.ChatMessagePartTypeText, Text: continuationSessionMarker + "\n" + text},
+			{Type: schema.ChatMessagePartTypeText, Text: "Please continue the conversation from where we left it off."},
+		},
+	}
+}
+
+func TestDedupContinuationUserMessages_KeepsLatest(t *testing.T) {
+	msgs := []adk.Message{
+		continuationUser("summary old"),
+		schema.UserMessage("real task"),
+		continuationUser("summary new"),
+	}
+	out, dropped := dedupContinuationUserMessages(msgs)
+	if dropped != 1 {
+		t.Fatalf("dropped=%d want 1", dropped)
+	}
+	if len(out) != 2 {
+		t.Fatalf("len=%d want 2", len(out))
+	}
+	if out[0].Role != schema.User || adkUserMessageText(out[0]) != "real task" {
+		t.Fatalf("first should remain real task, got %q", adkUserMessageText(out[0]))
+	}
+	if !strings.Contains(adkUserMessageText(out[1]), "summary new") {
+		t.Fatalf("latest continuation not kept: %q", adkUserMessageText(out[1]))
+	}
+}
+
+func TestDedupContinuationUserMessages_NoOpSingle(t *testing.T) {
+	msgs := []adk.Message{continuationUser("only"), schema.UserMessage("task")}
+	out, dropped := dedupContinuationUserMessages(msgs)
+	if dropped != 0 || len(out) != 2 {
+		t.Fatalf("unexpected change dropped=%d len=%d", dropped, len(out))
+	}
+}
+
+func TestContinuationUserDedupMiddleware(t *testing.T) {
+	mw := newContinuationUserDedupMiddleware(nil, "test")
+	state := &adk.ChatModelAgentState{Messages: []adk.Message{
+		continuationUser("old"),
+		continuationUser("new"),
+		schema.UserMessage("task"),
+	}}
+	_, out, err := mw.(*continuationUserDedupMiddleware).BeforeModelRewriteState(context.Background(), state, nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(out.Messages) != 2 {
+		t.Fatalf("want 2 messages after dedup, got %d", len(out.Messages))
+	}
+}

internal/multiagent/eino_adk_run_loop.go

@@ -18,6 +18,7 @@ import (
 	"cyberstrike-ai/internal/einomcp"
 	"cyberstrike-ai/internal/einoobserve"
 	"cyberstrike-ai/internal/openai"
+	"cyberstrike-ai/internal/security"
 
 	"github.com/cloudwego/eino/adk"
 	"github.com/cloudwego/eino/schema"
@@ -90,7 +91,7 @@ type einoADKRunLoopArgs struct {
 	FilesystemMonitorRecord einomcp.ExecutionRecorder
 	MCPExecutionBinder      *MCPExecutionBinder
 
-	// ToolInvokeNotify 与 einomcp.ToolsFromDefinitions 共享：run loop 在迭代前 Set，MCP 桥 Fire 以补全 tool_result。
+	// ToolInvokeNotify 与 einomcp.ToolsFromDefinitions 共享：run loop 在迭代前 Set，execute/MCP 桥 Fire 时立即推送 tool_result（ADK 晚到经 toolResultSent 去重）。
 	ToolInvokeNotify *einomcp.ToolInvokeNotifyHolder
 
 	DA adk.Agent
@@ -196,6 +197,16 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 		pendingByID[tc.ToolCallID] = tc
 		pendingQueueByAgent[tc.EinoAgent] = append(pendingQueueByAgent[tc.EinoAgent], tc.ToolCallID)
 	}
+	markPendingWithMonitor := func(tc toolCallPendingInfo) {
+		markPending(tc)
+		beginEinoADKFilesystemToolMonitor(
+			args.FilesystemMonitorAgent,
+			args.FilesystemMonitorRecord,
+			args.MCPExecutionBinder,
+			tc.ToolCallID,
+			tc.ToolName,
+		)
+	}
 	popNextPendingForAgent := func(agentName string) (toolCallPendingInfo, bool) {
 		pendingMu.Lock()
 		defer pendingMu.Unlock()
@@ -331,7 +342,7 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 			toolCallID = tid
 		}
 		recordPendingExecuteStdoutDup(toolName, content, isErr)
-		recordEinoADKFilesystemToolMonitor(args.FilesystemMonitorAgent, args.FilesystemMonitorRecord, toolName, toolCallID, runAccumulatedMsgs, content, isErr)
+		recordEinoADKFilesystemToolMonitor(args.FilesystemMonitorAgent, args.FilesystemMonitorRecord, args.MCPExecutionBinder, toolName, toolCallID, runAccumulatedMsgs, content, isErr)
 		if args.FilesystemMonitorAgent != nil && args.MCPExecutionBinder != nil {
 			if execID := args.MCPExecutionBinder.ExecutionID(toolCallID); execID != "" {
 				args.FilesystemMonitorAgent.UpdateMCPExecutionDisplayResult(execID, content)
@@ -341,8 +352,21 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 	}
 	if args.ToolInvokeNotify != nil {
 		args.ToolInvokeNotify.Set(func(toolCallID, toolName, einoAgent string, success bool, content string, invokeErr error) {
-			removePendingByID(strings.TrimSpace(toolCallID))
+			// Eino execute / MCP 桥在工具返回时 Fire；若 ADK schema.Tool 事件迟迟不到，此处立即推送
-			// tool_result 仅由下方 ADK schema.Tool 事件推送，正文与送入模型的上下文一致（含 reduction 截断）。
+			// tool_result 解除 UI「执行中」。tryEmitToolResultProgress 经 toolResultSent 去重，ADK 晚到不重复。
+			isErr := !success || invokeErr != nil
+			body := einoToolResultBody(content)
+			if einoToolResultIsError(toolName, content) {
+				isErr = true
+			}
+			if tail := friendlyEinoExecuteInvokeTail(invokeErr); tail != "" {
+				if body == "" {
+					body = tail
+				} else if !strings.Contains(body, tail) {
+					body = strings.TrimSpace(body) + "\n\n" + tail
+				}
+			}
+			tryEmitToolResultProgress(toolName, body, toolCallID, isErr, einoAgent)
 		})
 	}
 
@@ -383,6 +407,12 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 		}
 	}
 	runner := adk.NewRunner(ctx, runnerCfg)
+	startRunnerIter := func(runMsgs []adk.Message) *adk.AsyncIterator[*adk.AgentEvent] {
+		if checkPointID != "" {
+			return runner.Run(ctx, runMsgs, adk.WithCheckPointID(checkPointID))
+		}
+		return runner.Run(ctx, runMsgs)
+	}
 	var iter *adk.AsyncIterator[*adk.AgentEvent]
 	if cpStore != nil && checkPointID != "" {
 		if _, existed, getErr := cpStore.Get(ctx, checkPointID); getErr != nil {
@@ -422,12 +452,9 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 		}
 	}
 	if iter == nil {
-		if checkPointID != "" {
+		iter = startRunnerIter(msgs)
-			iter = runner.Run(ctx, msgs, adk.WithCheckPointID(checkPointID))
-		} else {
-			iter = runner.Run(ctx, msgs)
-		}
 	}
+	transientRetrier := newEinoTransientRunRetrier(einoTransientRunRetryPolicyFromArgs(args))
 	handleRunErr := func(runErr error) error {
 		if runErr == nil {
 			return nil
@@ -480,26 +507,67 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 		return runErr
 	}
 
-	// maybeRetryTransientRun：不在此层 runner.Run/Resume；由 handler 落库 + loadHistoryFromAgentTrace 分段续跑（同中断并继续）。
+	maybeRetryTransientRun := func(runErr error) (restarted bool, fatal error) {
-	maybeRetryTransientRun := func(runErr error) (retry bool, fatal error) {
+		if runErr == nil {
-		if runErr == nil || !isEinoTransientRunError(runErr) {
+			return false, nil
+		}
+		if !isEinoTransientRunError(runErr) {
 			return false, handleRunErr(runErr)
 		}
+		restarted, restartMsgs, ctxSource, backoff, retErr := transientRetrier.tryRetry(
+			ctx, runErr, args, baseMsgs, runAccumulatedMsgs, baseAccumulatedCount,
+		)
+		if retErr != nil {
+			flushAllPendingAsFailed(runErr)
 			if logger != nil {
-			logger.Warn("eino transient error, ending run segment for handler resume",
+				logger.Warn("eino transient retry exhausted",
+					zap.Error(retErr),
+					zap.String("orchestration", orchMode),
+					zap.Int("maxAttempts", transientRetrier.maxAttempts()))
+			}
+			return false, retErr
+		}
+		if !restarted {
+			return false, nil
+		}
+		attemptNo := transientRetrier.attempt()
+		maxAttempts := transientRetrier.maxAttempts()
+		if logger != nil {
+			logger.Warn("eino transient error, retrying after backoff",
 				zap.Error(runErr),
-				zap.String("orchestration", orchMode))
+				zap.String("orchestration", orchMode),
+				zap.Int("attempt", attemptNo),
+				zap.Int("maxAttempts", maxAttempts),
+				zap.Duration("backoff", backoff))
 		}
 		if progress != nil {
-			progress("eino_run_retry", "遇到临时错误（限流或网络波动），将保存上下文并重试…", map[string]interface{}{
+			progress("eino_run_retry", fmt.Sprintf("遇到临时错误（限流或网络波动），%d 秒后第 %d/%d 次重试…", int(backoff.Seconds()), attemptNo, maxAttempts), map[string]interface{}{
 				"conversationId": conversationID,
 				"source":         "eino",
 				"orchestration":  orchMode,
 				"error":          runErr.Error(),
-				"resumeKind":     "trace_segment",
+				"attempt":        attemptNo,
+				"maxAttempts":    maxAttempts,
+				"backoffSec":     int(backoff.Seconds()),
+			})
+			progress("eino_run_retry", "已恢复上下文，正在重试…", map[string]interface{}{
+				"conversationId": conversationID,
+				"source":         "eino",
+				"orchestration":  orchMode,
+				"attempt":        attemptNo,
+				"contextSource":  string(ctxSource),
 			})
 		}
-		return false, ErrTransientRetryContinue
+		msgs = restartMsgs
+		iter = startRunnerIter(msgs)
+		return true, nil
+	}
+
+	// 仅在退避重试后真正收到数据/完成一步时清零，避免重启后首个无错 ADK 事件误把计数打回 0。
+	confirmTransientRetryRecovery := func() {
+		if transientRetrier.attempt() > 0 {
+			transientRetrier.reset()
+		}
 	}
 
 	takePartial := func(runErr error) (*RunResult, error) {
@@ -514,10 +582,10 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 	}
 
 	for {
-		// 检测 context 取消（用户关闭浏览器、请求超时等），flush pending 工具状态避免 UI 卡在 "执行中"。
+		// iter.Next 可能长时间阻塞（工具执行、模型推理）；须与 ctx 联动，否则取消/超时无法及时 flush pending。
-		select {
+		ev, ok, iterCtxErr := nextAgentEventWithContext(ctx, iter)
-		case <-ctx.Done():
+		if iterCtxErr != nil {
-			flushAllPendingAsFailed(ctx.Err())
+			flushAllPendingAsFailed(iterCtxErr)
 			if progress != nil {
 				if isInterruptContinue(ctx) {
 					progress("progress", "已暂停当前输出，正在合并用户补充并继续…", map[string]interface{}{
@@ -526,17 +594,14 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 						"kind":           "interrupt_continue",
 					})
 				} else {
-					progress("error", "Request cancelled / 请求已取消", map[string]interface{}{
+					progress("error", iterCtxErr.Error(), map[string]interface{}{
 						"conversationId": conversationID,
 						"source":         "eino",
 					})
 				}
 			}
-			return takePartial(ctx.Err())
+			return takePartial(iterCtxErr)
-		default:
 		}
-
-		ev, ok := iter.Next()
 		if !ok {
 			// iter 结束并不总是“正常完成”：
 			// 当取消/超时发生在 iter.Next() 阻塞期间时，可能直接返回 !ok。
@@ -583,9 +648,13 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 			continue
 		}
 		if ev.Err != nil {
-			if _, retErr := maybeRetryTransientRun(ev.Err); retErr != nil {
+			restarted, retErr := maybeRetryTransientRun(ev.Err)
+			if retErr != nil {
 				return takePartial(retErr)
 			}
+			if restarted {
+				continue
+			}
 		}
 		if ev.AgentName != "" && progress != nil {
 			iterEinoAgent := orchestratorName
@@ -648,34 +717,9 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 
 		if mv.IsStreaming && mv.MessageStream != nil && mv.Role == schema.Tool {
 			toolName := strings.TrimSpace(mv.ToolName)
-			var toolBuf strings.Builder
+			content, streamToolCallID, toolStreamRecvErr := recvSchemaMessageStream(ctx, mv.MessageStream)
-			streamToolCallID := ""
+			isErr := einoToolResultIsError(toolName, content)
-			var toolStreamRecvErr error
+			content = einoToolResultBody(content)
-			for {
-				chunk, rerr := mv.MessageStream.Recv()
-				if errors.Is(rerr, io.EOF) {
-					break
-				}
-				if rerr != nil {
-					toolStreamRecvErr = rerr
-					break
-				}
-				if chunk == nil {
-					continue
-				}
-				if chunk.Content != "" {
-					toolBuf.WriteString(chunk.Content)
-				}
-				if tid := strings.TrimSpace(chunk.ToolCallID); tid != "" {
-					streamToolCallID = tid
-				}
-			}
-			content := toolBuf.String()
-			isErr := false
-			if strings.HasPrefix(content, einomcp.ToolErrorPrefix) {
-				isErr = true
-				content = strings.TrimPrefix(content, einomcp.ToolErrorPrefix)
-			}
 			if streamToolCallID != "" {
 				opts := []schema.ToolMessageOption{schema.WithToolName(toolName)}
 				runAccumulatedMsgs = append(runAccumulatedMsgs, schema.ToolMessage(content, streamToolCallID, opts...))
@@ -687,6 +731,9 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 					zap.String("agent", ev.AgentName),
 					zap.String("tool", toolName))
 			}
+			if toolStreamRecvErr == nil {
+				confirmTransientRetryRecovery()
+			}
 			continue
 		}
 
@@ -934,7 +981,7 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 			if merged := mergeStreamingToolCallFragments(toolStreamFragments); len(merged) > 0 {
 				lastToolChunk = mergeMessageToolCalls(&schema.Message{ToolCalls: merged})
 			}
-			tryEmitToolCallsOnce(lastToolChunk, ev.AgentName, orchestratorName, conversationID, orchMode, progress, toolEmitSeen, subAgentToolStep, mainAgentToolStep, markPending)
+			tryEmitToolCallsOnce(lastToolChunk, ev.AgentName, orchestratorName, conversationID, orchMode, progress, toolEmitSeen, subAgentToolStep, mainAgentToolStep, markPendingWithMonitor)
 			// 流式路径此前只把 tool_calls 推给进度 UI，未写入 runAccumulatedMsgs；落库后 loadHistory→RepairOrphan 会删掉全部 tool 结果，表现为「续跑/下轮失忆」。
 			if lastToolChunk != nil && len(lastToolChunk.ToolCalls) > 0 {
 				runAccumulatedMsgs = append(runAccumulatedMsgs, schema.AssistantMessage("", lastToolChunk.ToolCalls))
@@ -951,9 +998,15 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 						"einoRole":       einoRoleTag(ev.AgentName),
 					})
 				}
-				if _, retErr := maybeRetryTransientRun(streamRecvErr); retErr != nil {
+				restarted, retErr := maybeRetryTransientRun(streamRecvErr)
+				if retErr != nil {
 					return takePartial(retErr)
 				}
+				if restarted {
+					continue
+				}
+			} else {
+				confirmTransientRetryRecovery()
 			}
 			continue
 		}
@@ -963,7 +1016,7 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 			continue
 		}
 		runAccumulatedMsgs = append(runAccumulatedMsgs, msg)
-		tryEmitToolCallsOnce(mergeMessageToolCalls(msg), ev.AgentName, orchestratorName, conversationID, orchMode, progress, toolEmitSeen, subAgentToolStep, mainAgentToolStep, markPending)
+		tryEmitToolCallsOnce(mergeMessageToolCalls(msg), ev.AgentName, orchestratorName, conversationID, orchMode, progress, toolEmitSeen, subAgentToolStep, mainAgentToolStep, markPendingWithMonitor)
 
 		if mv.Role == schema.Assistant {
 			if progress != nil && strings.TrimSpace(msg.ReasoningContent) != "" {
@@ -1038,15 +1091,13 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 			}
 
 			content := msg.Content
-			isErr := false
+			isErr := einoToolResultIsError(toolName, content)
-			if strings.HasPrefix(content, einomcp.ToolErrorPrefix) {
+			content = einoToolResultBody(content)
-				isErr = true
-				content = strings.TrimPrefix(content, einomcp.ToolErrorPrefix)
-			}
 
 			toolCallID := strings.TrimSpace(msg.ToolCallID)
 			tryEmitToolResultProgress(toolName, content, toolCallID, isErr, ev.AgentName)
 		}
+		confirmTransientRetryRecovery()
 	}
 
 	mcpIDsMu.Lock()
@@ -1057,32 +1108,9 @@ func runEinoADKAgentLoop(ctx context.Context, args *einoADKRunLoopArgs, baseMsgs
 		orchMode, runAccumulatedMsgs, persistTraceSource(args, runAccumulatedMsgs),
 		lastAssistant, lastPlanExecuteExecutor, emptyHint, ids, false,
 	)
-	if shouldEinoEmptyResponseContinue(out, emptyHint, len(runAccumulatedMsgs), baseAccumulatedCount) {
-		if logger != nil {
-			logger.Info("eino empty response, ending run segment for handler resume",
-				zap.String("conversationId", conversationID),
-				zap.String("orchestration", orchMode),
-				zap.Int("traceMessages", len(runAccumulatedMsgs)))
-		}
-		if progress != nil {
-			progress("eino_empty_response_continue", "会话已结束但未产生助手正文，正在基于轨迹自动续跑…", map[string]interface{}{
-				"conversationId": conversationID,
-				"source":         "eino",
-				"resumeKind":     "trace_segment",
-			})
-		}
-		return out, ErrEmptyResponseContinue
-	}
 	return out, nil
 }
 
-func shouldEinoEmptyResponseContinue(out *RunResult, emptyHint string, accumulatedLen, baseCount int) bool {
-	if out == nil || accumulatedLen <= baseCount {
-		return false
-	}
-	return strings.TrimSpace(out.Response) == strings.TrimSpace(emptyHint)
-}
-
 func persistTraceSource(args *einoADKRunLoopArgs, fallback []adk.Message) []adk.Message {
 	if args != nil && args.ModelFacingTrace != nil {
 		if snap := args.ModelFacingTrace.Snapshot(); len(snap) > 0 {
@@ -1097,17 +1125,119 @@ func einoPartialRunLastOutputHint() string {
 		"[Run ended abnormally; continue from the trace above without repeating completed steps.]"
 }
 
-// friendlyEinoExecuteInvokeTail 将 Eino execute 等非 MCP 路径的结尾错误转成简短提示；其它情况保留原 error 文本。
+// friendlyEinoExecuteInvokeTail 将 Eino execute 超时/中断/流异常转为简短提示。
+// 命令非零退出（ExecuteExitError）已有 exec 对齐的正文，不再追加「执行未正常结束」。
 func friendlyEinoExecuteInvokeTail(invokeErr error) string {
 	if invokeErr == nil {
 		return ""
 	}
+	var exitErr *ExecuteExitError
+	if errors.As(invokeErr, &exitErr) {
+		return ""
+	}
 	if errors.Is(invokeErr, context.DeadlineExceeded) {
 		return einoExecuteTimeoutUserHint()
 	}
+	if errors.Is(invokeErr, context.Canceled) {
+		return ""
+	}
+	if strings.Contains(invokeErr.Error(), "shell inactivity timeout") {
+		return ""
+	}
 	return "[执行未正常结束] " + invokeErr.Error()
 }
 
+// einoToolResultIsError 统一判断 Eino 工具结果是否应标记为错误（与 MCP exec 的 IsError 对齐）。
+func einoToolResultIsError(toolName, content string) bool {
+	if strings.HasPrefix(content, einomcp.ToolErrorPrefix) {
+		return true
+	}
+	if strings.TrimSpace(toolName) == "execute" && security.IsCommandFailureResult(content) {
+		return true
+	}
+	return false
+}
+
+// einoToolResultBody 去掉工具错误前缀，返回展示/持久化正文。
+func einoToolResultBody(content string) string {
+	if strings.HasPrefix(content, einomcp.ToolErrorPrefix) {
+		return strings.TrimPrefix(content, einomcp.ToolErrorPrefix)
+	}
+	return content
+}
+
+// nextAgentEventWithContext 在 ctx 取消时不再无限阻塞于 iter.Next()（工具执行/模型推理期间常见）。
+func nextAgentEventWithContext(ctx context.Context, iter *adk.AsyncIterator[*adk.AgentEvent]) (ev *adk.AgentEvent, ok bool, ctxErr error) {
+	if iter == nil {
+		return nil, false, nil
+	}
+	type nextRes struct {
+		ev *adk.AgentEvent
+		ok bool
+	}
+	ch := make(chan nextRes, 1)
+	go func() {
+		e, o := iter.Next()
+		ch <- nextRes{e, o}
+	}()
+	select {
+	case <-ctx.Done():
+		return nil, false, ctx.Err()
+	case res := <-ch:
+		return res.ev, res.ok, nil
+	}
+}
+
+// recvSchemaMessageStream 消费 ADK Tool 流式结果；ctx 取消时立即返回，避免 amass 等无输出时永久阻塞。
+func recvSchemaMessageStream(ctx context.Context, stream *schema.StreamReader[*schema.Message]) (content, toolCallID string, recvErr error) {
+	if stream == nil {
+		return "", "", nil
+	}
+	type streamMsg struct {
+		chunk *schema.Message
+		err   error
+	}
+	recvCh := make(chan streamMsg, 8)
+	go func() {
+		defer close(recvCh)
+		for {
+			ch, rerr := stream.Recv()
+			recvCh <- streamMsg{chunk: ch, err: rerr}
+			if rerr != nil {
+				return
+			}
+		}
+	}()
+	var buf strings.Builder
+	for {
+		select {
+		case <-ctx.Done():
+			return buf.String(), toolCallID, ctx.Err()
+		case sm, open := <-recvCh:
+			if !open {
+				return buf.String(), toolCallID, nil
+			}
+			rerr := sm.err
+			if errors.Is(rerr, io.EOF) {
+				return buf.String(), toolCallID, nil
+			}
+			if rerr != nil {
+				return buf.String(), toolCallID, rerr
+			}
+			chunk := sm.chunk
+			if chunk == nil {
+				continue
+			}
+			if chunk.Content != "" {
+				buf.WriteString(chunk.Content)
+			}
+			if tid := strings.TrimSpace(chunk.ToolCallID); tid != "" {
+				toolCallID = tid
+			}
+		}
+	}
+}
+
 func buildEinoRunResultFromAccumulated(
 	orchMode string,
 	runAccumulatedMsgs []adk.Message,

internal/multiagent/eino_adk_run_loop_stream_test.go

@@ -0,0 +1,74 @@
+package multiagent
+
+import (
+	"context"
+	"errors"
+	"io"
+	"testing"
+	"time"
+
+	"github.com/cloudwego/eino/schema"
+)
+
+func TestRecvSchemaMessageStream_EOF(t *testing.T) {
+	sr, sw := schema.Pipe[*schema.Message](4)
+	_ = sw.Send(schema.ToolMessage("hello", "tc-1"), nil)
+	sw.Close()
+
+	content, tid, err := recvSchemaMessageStream(context.Background(), sr)
+	if err != nil {
+		t.Fatalf("unexpected err: %v", err)
+	}
+	if content != "hello" {
+		t.Fatalf("content=%q want hello", content)
+	}
+	if tid != "tc-1" {
+		t.Fatalf("toolCallID=%q want tc-1", tid)
+	}
+}
+
+func TestRecvSchemaMessageStream_ContextCancel(t *testing.T) {
+	sr, sw := schema.Pipe[*schema.Message](4)
+	t.Cleanup(func() { sw.Close() })
+
+	ctx, cancel := context.WithCancel(context.Background())
+	go func() {
+		time.Sleep(30 * time.Millisecond)
+		cancel()
+	}()
+
+	content, _, err := recvSchemaMessageStream(ctx, sr)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("want context.Canceled, got %v content=%q", err, content)
+	}
+}
+
+func TestRecvSchemaMessageStream_RecvError(t *testing.T) {
+	sr, sw := schema.Pipe[*schema.Message](4)
+	want := errors.New("stream broken")
+	_ = sw.Send(nil, want)
+	sw.Close()
+
+	_, _, err := recvSchemaMessageStream(context.Background(), sr)
+	if !errors.Is(err, want) {
+		t.Fatalf("want %v, got %v", want, err)
+	}
+}
+
+func TestRecvSchemaMessageStream_NilStream(t *testing.T) {
+	content, tid, err := recvSchemaMessageStream(context.Background(), nil)
+	if err != nil || content != "" || tid != "" {
+		t.Fatalf("nil stream: content=%q tid=%q err=%v", content, tid, err)
+	}
+}
+
+func TestRecvSchemaMessageStream_EOFViaEmptyRead(t *testing.T) {
+	sr, sw := schema.Pipe[*schema.Message](4)
+	_ = sw.Send(nil, io.EOF)
+	sw.Close()
+
+	_, _, err := recvSchemaMessageStream(context.Background(), sr)
+	if err != nil {
+		t.Fatalf("EOF should not surface as error, got %v", err)
+	}
+}

internal/multiagent/eino_chat_model_tail_middleware.go

@@ -0,0 +1,50 @@
+package multiagent
+
+import (
+	"github.com/cloudwego/eino/adk"
+	"go.uber.org/zap"
+)
+
+// einoChatModelTailConfig configures middleware appended after reduction/skill/plantask
+// and immediately before each ChatModel invocation pipeline completes.
+//
+// Order (best practice):
+//  1. system merge — accurate token count for summarization
+//  2. continuation user dedup — drop stale session-resume injections
+//  3. summarization
+//  4. orphan tool prune
+//  5. telemetry
+//  6. model-facing trace snapshot
+type einoChatModelTailConfig struct {
+	logger           *zap.Logger
+	phase            string
+	summarization    adk.ChatModelAgentMiddleware
+	modelName        string
+	conversationID   string
+	trace            *modelFacingTraceHolder
+	skipOrphanPruner bool
+	skipTelemetry    bool
+	skipTrace        bool
+}
+
+func appendEinoChatModelTailMiddlewares(handlers []adk.ChatModelAgentMiddleware, cfg einoChatModelTailConfig) []adk.ChatModelAgentMiddleware {
+	handlers = append(handlers, newSystemMessageNormalizerMiddleware(cfg.logger, cfg.phase))
+	handlers = append(handlers, newContinuationUserDedupMiddleware(cfg.logger, cfg.phase))
+	if cfg.summarization != nil {
+		handlers = append(handlers, cfg.summarization)
+	}
+	if !cfg.skipOrphanPruner {
+		handlers = append(handlers, newOrphanToolPrunerMiddleware(cfg.logger, cfg.phase))
+	}
+	if !cfg.skipTelemetry {
+		if teleMw := newEinoModelInputTelemetryMiddleware(cfg.logger, cfg.modelName, cfg.conversationID, cfg.phase); teleMw != nil {
+			handlers = append(handlers, teleMw)
+		}
+	}
+	if !cfg.skipTrace && cfg.trace != nil {
+		if capMw := newModelFacingTraceMiddleware(cfg.trace); capMw != nil {
+			handlers = append(handlers, capMw)
+		}
+	}
+	return handlers
+}

internal/multiagent/eino_empty_response_test.go

@@ -1,21 +0,0 @@
-package multiagent
-
-import "testing"
-
-func TestShouldEinoEmptyResponseContinue(t *testing.T) {
-	t.Parallel()
-	hint := "(empty hint)"
-	out := &RunResult{Response: hint}
-	if !shouldEinoEmptyResponseContinue(out, hint, 3, 1) {
-		t.Fatal("expected continue when response is empty hint and trace grew")
-	}
-	if shouldEinoEmptyResponseContinue(out, hint, 1, 1) {
-		t.Fatal("expected no continue when trace did not grow")
-	}
-	if shouldEinoEmptyResponseContinue(&RunResult{Response: "hello"}, hint, 3, 1) {
-		t.Fatal("expected no continue when response has content")
-	}
-	if shouldEinoEmptyResponseContinue(nil, hint, 3, 1) {
-		t.Fatal("expected no continue for nil result")
-	}
-}

internal/multiagent/eino_execute_failure_format_test.go

@@ -0,0 +1,114 @@
+package multiagent
+
+import (
+	"context"
+	"errors"
+	"io"
+	"strings"
+	"testing"
+
+	"cyberstrike-ai/internal/einomcp"
+	"cyberstrike-ai/internal/security"
+
+	"github.com/cloudwego/eino/adk/filesystem"
+	"github.com/cloudwego/eino/schema"
+)
+
+type mockStreamingShellExitFail struct {
+	output string
+	code   int
+}
+
+func (m *mockStreamingShellExitFail) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
+	outR, outW := schema.Pipe[*filesystem.ExecuteResponse](4)
+	go func() {
+		defer outW.Close()
+		if m.output != "" {
+			_ = outW.Send(&filesystem.ExecuteResponse{Output: m.output}, nil)
+		}
+		code := m.code
+		_ = outW.Send(&filesystem.ExecuteResponse{ExitCode: &code}, nil)
+	}()
+	return outR, nil
+}
+
+func TestEinoStreamingShellWrap_CommandFailureFormat(t *testing.T) {
+	inner := &mockStreamingShellExitFail{
+		output: "sudo: a password is required\n",
+		code:   1,
+	}
+	notify := einomcp.NewToolInvokeNotifyHolder()
+	var firedBody string
+	var firedSuccess bool
+	var firedErr error
+	notify.Set(func(toolCallID, toolName, einoAgent string, success bool, content string, invokeErr error) {
+		firedBody = content
+		firedSuccess = success
+		firedErr = invokeErr
+	})
+	wrap := &einoStreamingShellWrap{inner: inner, invokeNotify: notify}
+	sr, err := wrap.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: "sudo whoami"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	var stream strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+		if resp != nil {
+			stream.WriteString(resp.Output)
+		}
+	}
+
+	if firedSuccess {
+		t.Fatal("expected success=false")
+	}
+	var exitErr *ExecuteExitError
+	if !errors.As(firedErr, &exitErr) || exitErr.Code != 1 {
+		t.Fatalf("expected ExecuteExitError code 1, got %v", firedErr)
+	}
+	if !strings.HasPrefix(firedBody, einomcp.ToolErrorPrefix) {
+		t.Fatalf("missing tool error prefix: %q", firedBody)
+	}
+	body := strings.TrimPrefix(firedBody, einomcp.ToolErrorPrefix)
+	if body != security.FormatCommandFailureResult(1, "sudo: a password is required\n") {
+		t.Fatalf("fire body = %q", body)
+	}
+	if !strings.Contains(stream.String(), "sudo:") {
+		t.Fatalf("stream missing sudo output: %q", stream.String())
+	}
+	if strings.Contains(stream.String(), "command exited with non-zero") {
+		t.Fatalf("stream has legacy noise: %q", stream.String())
+	}
+	if strings.Contains(stream.String(), "执行未正常结束") {
+		t.Fatalf("stream has abnormal tail: %q", stream.String())
+	}
+	if !security.IsCommandFailureResult(stream.String()) {
+		t.Fatalf("stream missing failure status line: %q", stream.String())
+	}
+	if tail := friendlyEinoExecuteInvokeTail(firedErr); tail != "" {
+		t.Fatalf("unexpected invoke tail: %q", tail)
+	}
+	if !einoToolResultIsError("execute", firedBody) {
+		t.Fatal("expected isError for execute failure")
+	}
+}
+
+func TestFriendlyEinoExecuteInvokeTail(t *testing.T) {
+	if friendlyEinoExecuteInvokeTail(&ExecuteExitError{Code: 1}) != "" {
+		t.Fatal("exit error should not get abnormal tail")
+	}
+	if !strings.Contains(friendlyEinoExecuteInvokeTail(context.DeadlineExceeded), "Timed out") {
+		t.Fatal("deadline should get timeout hint")
+	}
+	if friendlyEinoExecuteInvokeTail(errors.New("broken pipe")) == "" {
+		t.Fatal("unexpected error should get tail")
+	}
+}

internal/multiagent/eino_execute_monitor.go

@@ -7,11 +7,25 @@ import (
 	"cyberstrike-ai/internal/einomcp"
 )
 
-// newEinoExecuteMonitorCallback 在 Eino filesystem execute 结束时写入 MCP 监控库并 recorder(executionId)，
+// newEinoExecuteMonitorCallbacks 在 Eino filesystem execute 开始/结束时写入 MCP 监控库并 recorder(executionId)，
-// 与 CallTool 路径一致，供助手消息展示「渗透测试详情」芯片。
+// 与 CallTool 路径一致，使监控页能展示「执行中」状态。
-func newEinoExecuteMonitorCallback(ag *agent.Agent, recorder einomcp.ExecutionRecorder) func(toolCallID, command, stdout string, success bool, invokeErr error) {
+func newEinoExecuteMonitorCallbacks(ag *agent.Agent, recorder einomcp.ExecutionRecorder) (
-	return func(toolCallID, command, stdout string, success bool, invokeErr error) {
+	begin func(toolCallID, command string) string,
-		if ag == nil || recorder == nil {
+	finish func(executionID, toolCallID, command, stdout string, success bool, invokeErr error),
+) {
+	begin = func(toolCallID, command string) string {
+		if ag == nil {
+			return ""
+		}
+		args := map[string]interface{}{"command": command}
+		id := ag.BeginLocalToolExecution("execute", args)
+		if id != "" && recorder != nil {
+			recorder(id, toolCallID)
+		}
+		return id
+	}
+	finish = func(executionID, toolCallID, command, stdout string, success bool, invokeErr error) {
+		if ag == nil {
 			return
 		}
 		var err error
@@ -23,9 +37,10 @@ func newEinoExecuteMonitorCallback(ag *agent.Agent, recorder einomcp.ExecutionRe
 			}
 		}
 		args := map[string]interface{}{"command": command}
-		id := ag.RecordLocalToolExecution("execute", args, stdout, err)
+		id := ag.FinishLocalToolExecution(executionID, "execute", args, stdout, err)
-		if id != "" {
+		if id != "" && recorder != nil && executionID == "" {
 			recorder(id, toolCallID)
 		}
 	}
+	return begin, finish
 }

internal/multiagent/eino_execute_streaming_wrap.go

@@ -6,9 +6,11 @@ import (
 	"fmt"
 	"io"
 	"strings"
+	"sync"
 	"time"
 
 	"cyberstrike-ai/internal/einomcp"
+	"cyberstrike-ai/internal/mcp"
 	"cyberstrike-ai/internal/security"
 
 	"github.com/cloudwego/eino/adk/filesystem"
@@ -49,7 +51,7 @@ func einoExecuteRecvErrIsToolTimeout(rerr error, tctx context.Context) bool {
 // 对「完全后台」命令自动开启 RunInBackendGround，与 local.runCmdInBackground 行为对齐。
 //
 // 使用 Pipe 将内层流转发给调用方：在 inner EOF 后、关闭 Pipe 前同步调用 ToolInvokeNotify.Fire，
-// 保证 run loop 在模型开始下一轮输出前已记录 execute 结果（用于 UI 与「重复助手复述」去重）。
+// run loop 收到 Fire 后立即推送 tool_result（toolResultSent 去重），避免 ADK Tool 事件迟到时 UI 卡在「执行中」。
 //
 // 若 inner 在校验阶段直接返回 error（未建立 reader），不会进入下方 goroutine，也必须 Fire；
 // 否则 pending tool_call 要等整轮 run 结束才被 force-close，与已展示的助手/工具软错误文案不同步。
@@ -61,8 +63,11 @@ type einoStreamingShellWrap struct {
 	outputChunk func(toolName, toolCallID, chunk string)
 	// toolTimeoutMinutes 与 agent.tool_timeout_minutes 对齐；>0 时对单次 execute 套用 context 超时（与 MCP 工具经 executeToolViaMCP 行为一致）。0 表示仅依赖上层 ctx（如整任务 10h 上限）。
 	toolTimeoutMinutes int
-	// recordMonitor 在 execute 流结束后写入 tool_executions 并 recorder(executionId)，使「渗透测试详情」与常规 MCP 一致。
+	// shellNoOutputTimeoutSec：无任何输出时的空闲秒数；0=关闭。
-	recordMonitor func(toolCallID, command, stdout string, success bool, invokeErr error)
+	shellNoOutputTimeoutSec int
+	// beginMonitor 在 execute 开始时写入 running 状态；finishMonitor 在流结束后更新为 completed/failed。
+	beginMonitor func(toolCallID, command string) string
+	finishMonitor func(executionID, toolCallID, command, stdout string, success bool, invokeErr error)
 }
 
 func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
@@ -74,43 +79,65 @@ func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *fi
 	}
 	req := *input
 	userCmd := strings.TrimSpace(req.Command)
+	tid := strings.TrimSpace(compose.GetToolCallID(ctx))
+	agentTag := strings.TrimSpace(w.einoAgentName)
 	if security.IsBackgroundShellCommand(req.Command) && !req.RunInBackendGround {
 		req.RunInBackendGround = true
 	}
-	req.Command = prependPythonUnbufferedEnv(req.Command)
+	req.Command = security.PrepareNonInteractiveShellCommand(prependPythonUnbufferedEnv(req.Command))
-	tid := strings.TrimSpace(compose.GetToolCallID(ctx))
+	convID := mcp.MCPConversationIDFromContext(ctx)
-	agentTag := strings.TrimSpace(w.einoAgentName)
+	execReg := mcp.EinoExecuteRunRegistryFromContext(ctx)
 
-	execCtx := ctx
+	var monitorExecID string
-	var execCancel context.CancelFunc
+	if w.beginMonitor != nil {
+		monitorExecID = w.beginMonitor(tid, userCmd)
+	}
+	if monitorExecID != "" && convID != "" {
+		if toolReg := mcp.ToolRunRegistryFromContext(ctx); toolReg != nil {
+			toolReg.RegisterRunningTool(convID, monitorExecID)
+		}
+	}
+	toolRunReg := mcp.ToolRunRegistryFromContext(ctx)
+
+	execCtx, execCancel := context.WithCancel(ctx)
+	var timeoutCancel context.CancelFunc
 	if w.toolTimeoutMinutes > 0 {
-		execCtx, execCancel = context.WithTimeout(ctx, time.Duration(w.toolTimeoutMinutes)*time.Minute)
+		execCtx, timeoutCancel = context.WithTimeout(execCtx, time.Duration(w.toolTimeoutMinutes)*time.Minute)
+	}
+	if execReg != nil && convID != "" {
+		execReg.RegisterActiveEinoExecute(convID, execCancel)
 	}
 
 	sr, err := w.inner.ExecuteStreaming(execCtx, &req)
 	if err != nil {
+		if timeoutCancel != nil {
+			timeoutCancel()
+		}
 		if execCancel != nil {
 			execCancel()
 		}
 		if einoExecuteRecvErrIsToolTimeout(err, execCtx) {
 			hint := "\n\n" + einoExecuteTimeoutUserHint() + "\n"
-			if w.recordMonitor != nil {
+			if w.finishMonitor != nil {
-				w.recordMonitor(tid, userCmd, hint, false, context.DeadlineExceeded)
+				w.finishMonitor(monitorExecID, tid, userCmd, hint, false, context.DeadlineExceeded)
 			}
 			if w.invokeNotify != nil && tid != "" {
 				w.invokeNotify.Fire(tid, "execute", agentTag, false, hint, context.DeadlineExceeded)
 			}
 			return schema.StreamReaderFromArray([]*filesystem.ExecuteResponse{{Output: hint}}), nil
 		}
-		if w.recordMonitor != nil {
+		if w.finishMonitor != nil {
-			w.recordMonitor(tid, userCmd, "", false, err)
+			w.finishMonitor(monitorExecID, tid, userCmd, "", false, err)
 		}
 		if w.invokeNotify != nil && tid != "" {
 			w.invokeNotify.Fire(tid, "execute", agentTag, false, "", err)
 		}
 		return nil, err
 	}
-	if sr == nil || w.invokeNotify == nil {
+	if sr == nil {
+		if timeoutCancel != nil {
+			timeoutCancel()
+		}
 		if execCancel != nil {
 			execCancel()
 		}
@@ -119,11 +146,35 @@ func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *fi
 
 	outR, outW := schema.Pipe[*filesystem.ExecuteResponse](32)
 
-	go func(inner *schema.StreamReader[*filesystem.ExecuteResponse], command string, cancel context.CancelFunc, tctx context.Context) {
+	go func(inner *schema.StreamReader[*filesystem.ExecuteResponse], command string, cancel context.CancelFunc, timeoutCleanup context.CancelFunc, tctx context.Context, conversationID string, reg mcp.EinoExecuteRunRegistry, toolReg mcp.ToolRunRegistry, execID string, toolCallID string, noOutputSec int) {
-		defer inner.Close()
+		var innerCloseOnce sync.Once
+		closeInner := func() {
+			innerCloseOnce.Do(func() { inner.Close() })
+		}
+		defer closeInner()
+		if timeoutCleanup != nil {
+			defer timeoutCleanup()
+		}
 		if cancel != nil {
 			defer cancel()
 		}
+		if reg != nil && conversationID != "" {
+			defer reg.UnregisterActiveEinoExecute(conversationID)
+		}
+		if toolReg != nil && conversationID != "" && execID != "" {
+			defer toolReg.UnregisterRunningTool(conversationID, execID)
+		}
+
+		// ctx 取消时关闭内层流，避免 amass 等长时间无换行输出时 Recv 永久阻塞。
+		stopWatch := make(chan struct{})
+		go func() {
+			select {
+			case <-tctx.Done():
+				closeInner()
+			case <-stopWatch:
+			}
+		}()
+		defer close(stopWatch)
 
 		var sb strings.Builder
 		success := true
@@ -131,46 +182,103 @@ func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *fi
 		exitCode := 0
 		hasExitCode := false
 
+		idleWatch := security.NewShellInactivityWatch(noOutputSec)
+		if idleWatch != nil {
+			defer idleWatch.Stop()
+		}
+
+		type execRecvMsg struct {
+			resp *filesystem.ExecuteResponse
+			err  error
+		}
+		recvCh := make(chan execRecvMsg, 1)
+		go func() {
 			for {
 				resp, rerr := inner.Recv()
+				recvCh <- execRecvMsg{resp: resp, err: rerr}
+				if rerr != nil {
+					return
+				}
+			}
+		}()
+
+		fireInactivityTimeout := func() {
+			success = false
+			invokeErr = fmt.Errorf("shell inactivity timeout (%ds)", idleWatch.Sec)
+			msg := security.ShellNoOutputTimeoutMessage(idleWatch.Sec)
+			_ = outW.Send(&filesystem.ExecuteResponse{Output: msg}, nil)
+			sb.WriteString(msg)
+			if w.outputChunk != nil && toolCallID != "" {
+				w.outputChunk("execute", toolCallID, msg)
+			}
+			if cancel != nil {
+				cancel()
+			}
+			closeInner()
+		}
+
+	recvLoop:
+		for {
+			var idleCh <-chan struct{}
+			if idleWatch != nil {
+				idleCh = idleWatch.Expired
+			}
+			select {
+			case <-idleCh:
+				fireInactivityTimeout()
+				break recvLoop
+			case msg := <-recvCh:
+				rerr := msg.err
+				resp := msg.resp
 				if errors.Is(rerr, io.EOF) {
-				break
+					break recvLoop
 				}
 				if rerr != nil {
 					success = false
 					invokeErr = rerr
-				// 单次 execute 超时须与 MCP 工具一致：写入工具结果尾标、继续迭代，不得向 ADK 流注入硬错误。
 					if einoExecuteRecvErrIsToolTimeout(rerr, tctx) {
 						invokeErr = context.DeadlineExceeded
-					break
+						break recvLoop
+					}
+					if errors.Is(rerr, context.Canceled) || (tctx != nil && errors.Is(tctx.Err(), context.Canceled)) {
+						invokeErr = context.Canceled
+						break recvLoop
 					}
 					_ = outW.Send(nil, rerr)
-				break
+					break recvLoop
 				}
 				if resp != nil {
 					if resp.ExitCode != nil {
 						hasExitCode = true
 						exitCode = *resp.ExitCode
+						continue
 					}
 					var appended string
 					if resp.Output != "" {
+						if security.IsLegacyShellExitNoise(resp.Output) {
+							continue
+						}
+						if idleWatch != nil {
+							idleWatch.Bump()
+						}
 						sb.WriteString(resp.Output)
 						appended = resp.Output
 					}
 					if w.outputChunk != nil && strings.TrimSpace(appended) != "" {
-					w.outputChunk("execute", tid, appended)
+						w.outputChunk("execute", toolCallID, appended)
 					}
 					if outW.Send(resp, nil) {
 						success = false
 						invokeErr = fmt.Errorf("execute stream closed by consumer")
-					break
+						break recvLoop
+					}
 				}
 			}
 		}
 
 		if success && hasExitCode && exitCode != 0 {
 			success = false
-			invokeErr = fmt.Errorf("execute exited with code %d", exitCode)
+			invokeErr = &ExecuteExitError{Code: exitCode}
 		}
 		// WithTimeout 触发后，子进程常被信号结束，local 侧多报 exit -1 / canceled，错误链里不一定带 DeadlineExceeded。
 		// 用执行所用 ctx 归一化，便于 UI 展示「超时」而非含糊的 -1。
@@ -178,6 +286,21 @@ func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *fi
 			success = false
 			invokeErr = context.DeadlineExceeded
 		}
+		// 用户「中断并继续」终止 execute：合并说明进工具结果（与 MCP CancelToolExecutionWithNote 一致）。
+		partialStreamed := sb.String()
+		var abortNote string
+		if reg != nil && conversationID != "" && (invokeErr != nil || errors.Is(tctx.Err(), context.Canceled)) {
+			if note := reg.TakeEinoExecuteAbortNote(conversationID); note != "" {
+				abortNote = note
+				merged := mcp.MergePartialToolOutputAndAbortNote(partialStreamed, note)
+				sb.Reset()
+				sb.WriteString(merged)
+				if invokeErr == nil {
+					success = false
+					invokeErr = context.Canceled
+				}
+			}
+		}
 		// ADK 从本 Pipe 拼出 tool 消息正文；仅 Notify 尾标不会进入模型上下文。超时句写入流，与 UI 一致。
 		if invokeErr != nil && errors.Is(invokeErr, context.DeadlineExceeded) {
 			hint := "\n\n" + einoExecuteTimeoutUserHint() + "\n"
@@ -187,12 +310,32 @@ func (w *einoStreamingShellWrap) ExecuteStreaming(ctx context.Context, input *fi
 			}
 			sb.WriteString(hint)
 		}
-		if w.recordMonitor != nil {
+		// 中断时循环内已逐行写入 stdout；此处只追加 USER INTERRUPT NOTE，避免整段输出重复。
-			w.recordMonitor(tid, command, sb.String(), success, invokeErr)
+		if invokeErr != nil && errors.Is(invokeErr, context.Canceled) && abortNote != "" {
+			if partialStreamed != "" {
+				_ = outW.Send(&filesystem.ExecuteResponse{Output: "\n\n" + mcp.AbortNoteBannerForModel + "\n" + abortNote}, nil)
+			} else if text := strings.TrimSpace(sb.String()); text != "" {
+				_ = outW.Send(&filesystem.ExecuteResponse{Output: text + "\n"}, nil)
+			}
+		}
+		rawOutput := sb.String()
+		fireBody := rawOutput
+		if !success && hasExitCode && exitCode != 0 {
+			statusLine := security.ExecuteFailureStatusLine(exitCode)
+			if !strings.Contains(rawOutput, "命令执行失败:") {
+				_ = outW.Send(&filesystem.ExecuteResponse{Output: statusLine}, nil)
+				sb.WriteString(statusLine)
+			}
+			fireBody = einomcp.ToolErrorPrefix + security.FormatCommandFailureResult(exitCode, rawOutput)
+		}
+		if w.finishMonitor != nil {
+			w.finishMonitor(execID, toolCallID, command, sb.String(), success, invokeErr)
+		}
+		if w.invokeNotify != nil {
+			w.invokeNotify.Fire(toolCallID, "execute", agentTag, success, fireBody, invokeErr)
 		}
-		w.invokeNotify.Fire(tid, "execute", agentTag, success, sb.String(), invokeErr)
 		outW.Close()
-	}(sr, userCmd, execCancel, execCtx)
+	}(sr, userCmd, execCancel, timeoutCancel, execCtx, convID, execReg, toolRunReg, monitorExecID, tid, w.shellNoOutputTimeoutSec)
 
 	return outR, nil
 }

internal/multiagent/eino_execute_streaming_wrap_test.go

@@ -9,6 +9,7 @@ import (
 	"time"
 
 	"cyberstrike-ai/internal/einomcp"
+	"cyberstrike-ai/internal/mcp"
 
 	"github.com/cloudwego/eino/adk/filesystem"
 	"github.com/cloudwego/eino/schema"
@@ -18,9 +19,15 @@ type mockStreamingShell struct {
 	immediateErr error
 	recvErr      error
 	output       string
+	called       bool
+	lastCommand  string
 }
 
 func (m *mockStreamingShell) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
+	m.called = true
+	if input != nil {
+		m.lastCommand = input.Command
+	}
 	if m.immediateErr != nil {
 		return nil, m.immediateErr
 	}
@@ -37,6 +44,135 @@ func (m *mockStreamingShell) ExecuteStreaming(ctx context.Context, input *filesy
 	return outR, nil
 }
 
+func TestEinoStreamingShellWrap_PreparesNonInteractiveCommand(t *testing.T) {
+	inner := &mockStreamingShell{output: "ok\n"}
+	wrap := &einoStreamingShellWrap{inner: inner}
+	sr, err := wrap.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: "echo ok"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+	for {
+		_, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+	}
+	if !strings.Contains(inner.lastCommand, "exec </dev/null") {
+		t.Fatalf("missing stdin redirect in inner command: %q", inner.lastCommand)
+	}
+	if !strings.Contains(inner.lastCommand, "GIT_PAGER=cat") {
+		t.Fatalf("missing pager export in inner command: %q", inner.lastCommand)
+	}
+	if !strings.Contains(inner.lastCommand, "PYTHONUNBUFFERED=1") {
+		t.Fatalf("missing python unbuffer in inner command: %q", inner.lastCommand)
+	}
+}
+
+func TestEinoStreamingShellWrap_NoOutputTimeout(t *testing.T) {
+	inner := &mockStreamingShellHanging{}
+	notify := einomcp.NewToolInvokeNotifyHolder()
+	var fired string
+	notify.Set(func(toolCallID, toolName, einoAgent string, success bool, content string, invokeErr error) {
+		fired = content
+	})
+	wrap := &einoStreamingShellWrap{
+		inner:                   inner,
+		invokeNotify:            notify,
+		shellNoOutputTimeoutSec: 1,
+	}
+	sr, err := wrap.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: "sudo whoami"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+		if resp != nil {
+			got.WriteString(resp.Output)
+		}
+	}
+	if !inner.called {
+		t.Fatal("inner shell should run (no command blacklist)")
+	}
+	out := got.String()
+	if !strings.Contains(out, "没有新的输出") && !strings.Contains(out, "no new output") {
+		t.Fatalf("expected inactivity timeout message, got: %q notify=%q", out, fired)
+	}
+}
+
+type mockStreamingShellPartialThenHang struct {
+	called bool
+}
+
+func (m *mockStreamingShellPartialThenHang) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
+	m.called = true
+	outR, outW := schema.Pipe[*filesystem.ExecuteResponse](4)
+	go func() {
+		_ = outW.Send(&filesystem.ExecuteResponse{Output: "[sudo] password:\n"}, nil)
+		<-ctx.Done()
+		outW.Close()
+	}()
+	return outR, nil
+}
+
+func TestEinoStreamingShellWrap_InactivityAfterPartialOutput(t *testing.T) {
+	inner := &mockStreamingShellPartialThenHang{}
+	wrap := &einoStreamingShellWrap{
+		inner:                   inner,
+		shellNoOutputTimeoutSec: 1,
+	}
+	start := time.Now()
+	sr, err := wrap.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: "sudo whoami"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+		if resp != nil {
+			got.WriteString(resp.Output)
+		}
+	}
+	if time.Since(start) > 5*time.Second {
+		t.Fatalf("expected inactivity timeout ~1s, took %v", time.Since(start))
+	}
+	if !strings.Contains(got.String(), "没有新的输出") && !strings.Contains(got.String(), "no new output") {
+		t.Fatalf("expected inactivity message, got: %q", got.String())
+	}
+}
+
+type mockStreamingShellHanging struct {
+	called bool
+}
+
+func (m *mockStreamingShellHanging) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
+	m.called = true
+	outR, outW := schema.Pipe[*filesystem.ExecuteResponse](4)
+	go func() {
+		<-ctx.Done()
+		outW.Close()
+	}()
+	return outR, nil
+}
+
 func TestEinoExecuteRecvErrIsToolTimeout(t *testing.T) {
 	tctx, cancel := context.WithTimeout(context.Background(), time.Millisecond)
 	defer cancel()
@@ -122,6 +258,94 @@ func TestEinoStreamingShellWrap_ToolTimeoutRecvErrIsSoft(t *testing.T) {
 	}
 }
 
+func TestEinoStreamingShellWrap_CapturesOutputWithToolTimeout(t *testing.T) {
+	inner := &mockStreamingShell{output: "100\n"}
+	notify := einomcp.NewToolInvokeNotifyHolder()
+	var firedContent string
+	notify.Set(func(toolCallID, toolName, einoAgent string, success bool, content string, invokeErr error) {
+		firedContent = content
+	})
+	wrap := &einoStreamingShellWrap{
+		inner:              inner,
+		invokeNotify:       notify,
+		toolTimeoutMinutes: 60,
+	}
+	sr, err := wrap.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: "echo 100"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("unexpected stream error: %v", rerr)
+		}
+		if resp != nil && resp.Output != "" {
+			got.WriteString(resp.Output)
+		}
+	}
+	if !strings.Contains(got.String(), "100") {
+		t.Fatalf("stream output = %q, want contains 100", got.String())
+	}
+	if !strings.Contains(firedContent, "100") {
+		t.Fatalf("notify content = %q, want contains 100", firedContent)
+	}
+}
+
+func TestEinoStreamingShellWrap_AbortNoteDoesNotDuplicateStreamedOutput(t *testing.T) {
+	inner := &mockStreamingShell{output: "line1\nline2\n", recvErr: context.Canceled}
+	notify := einomcp.NewToolInvokeNotifyHolder()
+	wrap := &einoStreamingShellWrap{
+		inner:        inner,
+		invokeNotify: notify,
+	}
+	reg := &abortNoteTestRegistry{note: "改成20次"}
+	ctx := mcp.WithEinoExecuteRunRegistry(
+		mcp.WithMCPConversationID(context.Background(), "conv-abort-dup"),
+		reg,
+	)
+	sr, err := wrap.ExecuteStreaming(ctx, &filesystem.ExecuteRequest{Command: "ping -c 10 baidu.com"})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("unexpected stream error: %v", rerr)
+		}
+		if resp != nil && resp.Output != "" {
+			got.WriteString(resp.Output)
+		}
+	}
+	out := got.String()
+	if strings.Count(out, "line1") != 1 || strings.Count(out, "line2") != 1 {
+		t.Fatalf("stream duplicated stdout: %q", out)
+	}
+	if !strings.Contains(out, "改成20次") {
+		t.Fatalf("stream missing abort note: %q", out)
+	}
+}
+
+type abortNoteTestRegistry struct {
+	note string
+}
+
+func (r *abortNoteTestRegistry) RegisterActiveEinoExecute(string, context.CancelFunc) {}
+func (r *abortNoteTestRegistry) UnregisterActiveEinoExecute(string)                   {}
+func (r *abortNoteTestRegistry) AbortActiveEinoExecute(string, string) bool           { return false }
+func (r *abortNoteTestRegistry) TakeEinoExecuteAbortNote(string) string               { return r.note }
+
 func TestEinoStreamingShellWrap_NonTimeoutRecvErrStillHard(t *testing.T) {
 	inner := &mockStreamingShell{recvErr: errors.New("broken pipe")}
 	wrap := &einoStreamingShellWrap{inner: inner}

internal/multiagent/eino_filesystem_tool_monitor.go

@@ -63,10 +63,43 @@ func toolCallArgsFromAccumulated(msgs []adk.Message, toolCallID, expectToolName
 	return map[string]interface{}{}
 }
 
+// beginEinoADKFilesystemToolMonitor 在 Eino ADK filesystem 工具开始调用时写入 running 状态。
+func beginEinoADKFilesystemToolMonitor(
+	ag *agent.Agent,
+	rec einomcp.ExecutionRecorder,
+	binder *MCPExecutionBinder,
+	toolCallID, toolName string,
+) {
+	if ag == nil || rec == nil {
+		return
+	}
+	name := strings.TrimSpace(toolName)
+	if name == "" || strings.EqualFold(name, "execute") {
+		return
+	}
+	if !isBuiltinEinoADKFilesystemToolName(name) {
+		return
+	}
+	tid := strings.TrimSpace(toolCallID)
+	if tid == "" {
+		return
+	}
+	storedName := "eino_fs::" + strings.ToLower(name)
+	id := ag.BeginLocalToolExecution(storedName, map[string]interface{}{})
+	if id == "" {
+		return
+	}
+	rec(id, tid)
+	if binder != nil {
+		binder.Bind(tid, id)
+	}
+}
+
 // recordEinoADKFilesystemToolMonitor 将 Eino ADK filesystem 中间件工具结果写入 MCP 监控（与 execute / MCP 桥芯片一致）。
 func recordEinoADKFilesystemToolMonitor(
 	ag *agent.Agent,
 	rec einomcp.ExecutionRecorder,
+	binder *MCPExecutionBinder,
 	toolName string,
 	toolCallID string,
 	msgs []adk.Message,
@@ -94,8 +127,12 @@ func recordEinoADKFilesystemToolMonitor(
 			invErr = errors.New(t)
 		}
 	}
-	id := ag.RecordLocalToolExecution(storedName, args, resultText, invErr)
+	execID := ""
-	if id != "" {
+	if binder != nil {
+		execID = binder.ExecutionID(toolCallID)
+	}
+	id := ag.FinishLocalToolExecution(execID, storedName, args, resultText, invErr)
+	if id != "" && execID == "" {
 		rec(id, toolCallID)
 	}
 }

internal/multiagent/eino_middleware.go

@@ -243,17 +243,14 @@ func prependEinoMiddlewares(
 	return outTools, extraHandlers, toolSearchActive, nil
 }
 
-func deepExtrasFromConfig(ma *config.MultiAgentConfig) (outputKey string, retry *adk.ModelRetryConfig, taskDesc func(context.Context, []adk.Agent) (string, error)) {
+func deepExtrasFromConfig(ma *config.MultiAgentConfig) (outputKey string, taskDesc func(context.Context, []adk.Agent) (string, error)) {
 	if ma == nil {
-		return "", nil, nil
+		return "", nil
 	}
 	mw := ma.EinoMiddleware
 	if k := strings.TrimSpace(mw.DeepOutputKey); k != "" {
 		outputKey = k
 	}
-	if mw.DeepModelRetryMaxRetries > 0 {
-		retry = &adk.ModelRetryConfig{MaxRetries: mw.DeepModelRetryMaxRetries}
-	}
 	prefix := strings.TrimSpace(mw.TaskToolDescriptionPrefix)
 	if prefix != "" {
 		taskDesc = func(ctx context.Context, agents []adk.Agent) (string, error) {
@@ -274,5 +271,5 @@ func deepExtrasFromConfig(ma *config.MultiAgentConfig) (outputKey string, retry
 			return prefix + "\n可用子代理（按名称 transfer / task 调用）：" + strings.Join(names, "、"), nil
 		}
 	}
-	return outputKey, retry, taskDesc
+	return outputKey, taskDesc
 }

internal/multiagent/eino_orchestration.go

@@ -94,24 +94,20 @@ func NewPlanExecuteRoot(ctx context.Context, a *PlanExecuteRootArgs) (adk.Resuma
 	if a.SkillMiddleware != nil {
 		execHandlers = append(execHandlers, a.SkillMiddleware)
 	}
-	// 4. summarization（最后，与 Deep/Supervisor 一致）
+	// 4. pre-summarization normalize + continuation dedup, then summarization (与 Deep/Supervisor 一致)
 	if a.AppCfg != nil {
 		sumMw, sumErr := newEinoSummarizationMiddleware(ctx, a.ExecModel, a.AppCfg, a.MwCfg, a.ConversationID, a.DB, a.ProjectID, a.Logger)
 		if sumErr != nil {
 			return nil, fmt.Errorf("plan_execute executor summarization: %w", sumErr)
 		}
-		execHandlers = append(execHandlers, sumMw)
+		execHandlers = appendEinoChatModelTailMiddlewares(execHandlers, einoChatModelTailConfig{
-	}
+			logger:         a.Logger,
-	// 5. 孤儿 tool 消息兜底：必须挂在所有改写历史中间件（summarization/reduction/skill）之后、
+			phase:          "plan_execute_executor",
-	//    telemetry 之前，保证送入 ChatModel 的消息序列 tool_call ↔ tool_result 配对完整。
+			summarization:  sumMw,
-	execHandlers = append(execHandlers, newOrphanToolPrunerMiddleware(a.Logger, "plan_execute_executor"))
+			modelName:      a.ModelName,
-	if teleMw := newEinoModelInputTelemetryMiddleware(a.Logger, a.ModelName, a.ConversationID, "plan_execute_executor"); teleMw != nil {
+			conversationID: a.ConversationID,
-		execHandlers = append(execHandlers, teleMw)
+			trace:          a.ModelFacingTrace,
-	}
+		})
-	if a.ModelFacingTrace != nil {
-		if capMw := newModelFacingTraceMiddleware(a.ModelFacingTrace); capMw != nil {
-			execHandlers = append(execHandlers, capMw)
-		}
 	}
 	executor, err := newPlanExecuteExecutor(ctx, &planexecute.ExecutorConfig{
 		Model:         a.ExecModel,

internal/multiagent/eino_single_runner.go

@@ -81,7 +81,7 @@ func RunEinoSingleChatModelAgent(
 	}
 
 	toolInvokeNotify := einomcp.NewToolInvokeNotifyHolder()
-	einoExecMonitor := newEinoExecuteMonitorCallback(ag, recorder)
+	einoExecBegin, einoExecFinish := newEinoExecuteMonitorCallbacks(ag, recorder)
 	mainDefs := ag.ToolsForRole(roleTools)
 	mainTools, err := einomcp.ToolsFromDefinitions(ag, holder, mainDefs, recorder, nil, toolInvokeNotify, einoSingleAgentName)
 	if err != nil {
@@ -136,7 +136,7 @@ func RunEinoSingleChatModelAgent(
 	}
 	if einoSkillMW != nil {
 		if einoFSTools && einoLoc != nil {
-			fsMw, fsErr := subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, einoSingleAgentName, einoExecMonitor, agentToolTimeoutMinutes(appCfg), nil)
+			fsMw, fsErr := subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, einoSingleAgentName, einoExecBegin, einoExecFinish, agentToolTimeoutMinutes(appCfg), agentShellNoOutputTimeoutSeconds(appCfg), nil)
 			if fsErr != nil {
 				return nil, fmt.Errorf("eino single filesystem 中间件: %w", fsErr)
 			}
@@ -144,13 +144,14 @@ func RunEinoSingleChatModelAgent(
 		}
 		handlers = append(handlers, einoSkillMW)
 	}
-	handlers = append(handlers, mainSumMw)
+	handlers = appendEinoChatModelTailMiddlewares(handlers, einoChatModelTailConfig{
-	if teleMw := newEinoModelInputTelemetryMiddleware(logger, appCfg.OpenAI.Model, conversationID, "eino_single"); teleMw != nil {
+		logger:         logger,
-		handlers = append(handlers, teleMw)
+		phase:          "eino_single",
-	}
+		summarization:  mainSumMw,
-	if capMw := newModelFacingTraceMiddleware(modelFacingTrace); capMw != nil {
+		modelName:      appCfg.OpenAI.Model,
-		handlers = append(handlers, capMw)
+		conversationID: conversationID,
-	}
+		trace:          modelFacingTrace,
+	})
 
 	maxIter := agentMaxIterations(appCfg)
 
@@ -188,13 +189,10 @@ func RunEinoSingleChatModelAgent(
 		MaxIterations: maxIter,
 		Handlers:      handlers,
 	}
-	outKey, modelRetry, _ := deepExtrasFromConfig(ma)
+	outKey, _ := deepExtrasFromConfig(ma)
 	if outKey != "" {
 		chatCfg.OutputKey = outKey
 	}
-	if modelRetry != nil {
-		chatCfg.ModelRetryConfig = modelRetry
-	}
 
 	chatAgent, err := adk.NewChatModelAgent(ctx, chatCfg)
 	if err != nil {

internal/multiagent/eino_skills.go

@@ -9,6 +9,7 @@ import (
 
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/einomcp"
+	"cyberstrike-ai/internal/security"
 
 	localbk "github.com/cloudwego/eino-ext/adk/backend/local"
 	"github.com/cloudwego/eino/adk"
@@ -81,8 +82,10 @@ func subAgentFilesystemMiddleware(
 	loc *localbk.Local,
 	invokeNotify *einomcp.ToolInvokeNotifyHolder,
 	einoAgentName string,
-	recordMonitor func(toolCallID, command, stdout string, success bool, invokeErr error),
+	beginMonitor func(toolCallID, command string) string,
+	finishMonitor func(executionID, toolCallID, command, stdout string, success bool, invokeErr error),
 	toolTimeoutMinutes int,
+	shellNoOutputTimeoutSec int,
 	outputChunk func(toolName, toolCallID, chunk string),
 ) (adk.ChatModelAgentMiddleware, error) {
 	if loc == nil {
@@ -91,12 +94,14 @@ func subAgentFilesystemMiddleware(
 	return filesystem.New(ctx, &filesystem.MiddlewareConfig{
 		Backend: loc,
 		StreamingShell: &einoStreamingShellWrap{
-			inner:              loc,
+			inner:                   security.NewEinoStreamingShell(),
 			invokeNotify:            invokeNotify,
 			einoAgentName:           strings.TrimSpace(einoAgentName),
 			outputChunk:             outputChunk,
-			recordMonitor:      recordMonitor,
+			beginMonitor:            beginMonitor,
+			finishMonitor:           finishMonitor,
 			toolTimeoutMinutes:      toolTimeoutMinutes,
+			shellNoOutputTimeoutSec: shellNoOutputTimeoutSec,
 		},
 	})
 }
@@ -108,3 +113,18 @@ func agentToolTimeoutMinutes(cfg *config.Config) int {
 	}
 	return cfg.Agent.ToolTimeoutMinutes
 }
+
+// agentShellNoOutputTimeoutSeconds：0=默认 300s（5 分钟）；-1=关闭；>0=自定义秒数。
+func agentShellNoOutputTimeoutSeconds(cfg *config.Config) int {
+	if cfg == nil {
+		return 300
+	}
+	v := cfg.Agent.ShellNoOutputTimeoutSeconds
+	if v < 0 {
+		return 0
+	}
+	if v == 0 {
+		return 300
+	}
+	return v
+}

internal/multiagent/eino_summarize.go

@@ -22,8 +22,6 @@ import (
 	"go.uber.org/zap"
 )
 
-const defaultSummarizationRetryMax = 3
-
 // einoSummarizeUserInstruction：压缩历史时保留渗透测试关键信息。
 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。
 
@@ -97,10 +95,8 @@ func newEinoSummarizationMiddleware(
 		}
 	}
 
-	retryMax := defaultSummarizationRetryMax
+	retryPolicy := einoTransientRunRetryPolicyFromMW(mwCfg)
-	if mwCfg != nil && mwCfg.SummarizationRetryMaxAttempts > 0 {
+	retryMax := retryPolicy.maxAttempts
-		retryMax = mwCfg.SummarizationRetryMaxAttempts
-	}
 
 	// ModelOptions apply only to summarization Generate (same ChatModel instance as the agent).
 	// Strip thinking/reasoning on this call path; mark requests for empty-choices diagnostics.
@@ -137,13 +133,14 @@ func newEinoSummarizationMiddleware(
 		Retry: &summarization.RetryConfig{
 			MaxRetries: &retryMax,
 			ShouldRetry: func(_ context.Context, _ adk.Message, err error) bool {
-				if err != nil && logger != nil {
+				retry := isEinoTransientRunError(err)
-					logger.Warn("eino summarization generate attempt failed, will retry if attempts remain",
+				if retry && logger != nil {
+					logger.Warn("eino summarization generate transient error, will retry if attempts remain",
 						zap.Error(err),
 						zap.Int("max_retries", retryMax),
 					)
 				}
-				return err != nil
+				return retry
 			},
 		},
 		Finalize: func(ctx context.Context, originalMessages []adk.Message, summary adk.Message) ([]adk.Message, error) {
@@ -260,17 +257,19 @@ func summarizeFinalizeWithRecentAssistantToolTrail(
 		nonSystem = append(nonSystem, msg)
 	}
 
+	mergedSystem := mergeCollectedSystemMessages(systemMsgs)
+
 	if recentTrailTokenBudget <= 0 || len(nonSystem) == 0 {
-		out := make([]adk.Message, 0, len(systemMsgs)+1)
+		out := make([]adk.Message, 0, len(mergedSystem)+1)
-		out = append(out, systemMsgs...)
+		out = append(out, mergedSystem...)
 		out = append(out, summary)
 		return out, nil
 	}
 
 	rounds := splitMessagesIntoRounds(nonSystem)
 	if len(rounds) == 0 {
-		out := make([]adk.Message, 0, len(systemMsgs)+1)
+		out := make([]adk.Message, 0, len(mergedSystem)+1)
-		out = append(out, systemMsgs...)
+		out = append(out, mergedSystem...)
 		out = append(out, summary)
 		return out, nil
 	}
@@ -322,8 +321,8 @@ func summarizeFinalizeWithRecentAssistantToolTrail(
 		selectedMsgs = append(selectedMsgs, selectedRoundsReverse[i].messages...)
 	}
 
-	out := make([]adk.Message, 0, len(systemMsgs)+1+len(selectedMsgs))
+	out := make([]adk.Message, 0, len(mergedSystem)+1+len(selectedMsgs))
-	out = append(out, systemMsgs...)
+	out = append(out, mergedSystem...)
 	out = append(out, summary)
 	out = append(out, selectedMsgs...)
 	return out, nil

internal/multiagent/eino_summarize_test.go

@@ -192,8 +192,8 @@ func TestSummarizeFinalize_KeepsToolRoundIntact(t *testing.T) {
 	if len(out) < 2 {
 		t.Fatalf("output too short: %d", len(out))
 	}
-	if out[0] != sys {
+	if out[0].Role != schema.System || out[0].Content != "sys" {
-		t.Fatalf("first message must be system")
+		t.Fatalf("first message must be system sys, got %s: %q", out[0].Role, out[0].Content)
 	}
 	if out[1] != summary {
 		t.Fatalf("second message must be summary")
@@ -293,12 +293,12 @@ func TestSummarizeFinalize_BudgetZeroFallsBackToSummaryOnly(t *testing.T) {
 	if err != nil {
 		t.Fatalf("unexpected error: %v", err)
 	}
-	if len(out) != 2 || out[0] != sys || out[1] != summary {
+	if len(out) != 2 || out[0].Role != schema.System || out[0].Content != "sys" || out[1] != summary {
 		t.Fatalf("budget=0 must yield [system, summary] only, got %+v", out)
 	}
 }
 
-func TestSummarizeFinalize_PreservesAllSystemMessages(t *testing.T) {
+func TestSummarizeFinalize_MergesSystemMessages(t *testing.T) {
 	sys1 := schema.SystemMessage("sys1")
 	sys2 := schema.SystemMessage("sys2")
 	summary := schema.AssistantMessage("s", nil)
@@ -321,10 +321,13 @@ func TestSummarizeFinalize_PreservesAllSystemMessages(t *testing.T) {
 	for _, m := range out {
 		if m != nil && m.Role == schema.System {
 			systemCount++
+			if got := m.Content; got != "sys1\n\nsys2" {
+				t.Fatalf("unexpected merged system content: %q", got)
 			}
 		}
-	if systemCount != 2 {
+	}
-		t.Fatalf("want 2 system messages retained, got %d", systemCount)
+	if systemCount != 1 {
+		t.Fatalf("want 1 merged system message, got %d", systemCount)
 	}
 }
 
@@ -378,6 +381,12 @@ func TestWriteSummarizationTranscript(t *testing.T) {
 	if !strings.Contains(text, "tool_calls:") || !strings.Contains(text, "nmap output") {
 		t.Fatalf("missing tool round: %q", text)
 	}
+	if !strings.Contains(text, `"name":"stub_tool"`) || !strings.Contains(text, `"arguments":"{}"`) {
+		t.Fatalf("missing tool name/arguments: %q", text)
+	}
+	if strings.Contains(text, "tool_call_id") || strings.Contains(text, `"id":"tc1"`) {
+		t.Fatalf("transcript should omit tool_call_id: %q", text)
+	}
 }
 
 func TestSanitizeSystemContentForTranscript_BestPractice(t *testing.T) {

internal/multiagent/eino_summarize_transcript.go

@@ -23,6 +23,11 @@ const (
 	transcriptSkillsSystemMarker     = "# Skills System"
 )
 
+type transcriptToolCall struct {
+	Name      string `json:"name"`
+	Arguments string `json:"arguments"`
+}
+
 // formatSummarizationTranscript renders pre-compaction messages for transcript.txt.
 // Best practice: keep full user/assistant/tool turns; slim system to dynamic blocks only.
 func formatSummarizationTranscript(msgs []adk.Message) string {
@@ -138,15 +143,21 @@ func appendTranscriptMessage(sb *strings.Builder, msg adk.Message) {
 		}
 	}
 	if len(msg.ToolCalls) > 0 {
-		if b, err := sonic.Marshal(msg.ToolCalls); err == nil {
+		if b, err := sonic.Marshal(formatTranscriptToolCalls(msg.ToolCalls)); err == nil {
 			sb.WriteString("tool_calls: ")
 			sb.Write(b)
 			sb.WriteByte('\n')
 		}
 	}
-	if msg.ToolCallID != "" {
-		sb.WriteString("tool_call_id: ")
-		sb.WriteString(msg.ToolCallID)
-		sb.WriteByte('\n')
 }
+
+func formatTranscriptToolCalls(calls []schema.ToolCall) []transcriptToolCall {
+	out := make([]transcriptToolCall, 0, len(calls))
+	for _, tc := range calls {
+		out = append(out, transcriptToolCall{
+			Name:      tc.Function.Name,
+			Arguments: tc.Function.Arguments,
+		})
+	}
+	return out
 }

internal/multiagent/eino_tool_name_injection.go

@@ -46,6 +46,10 @@ func injectToolNamesOnlyInstruction(ctx context.Context, instruction string, too
 		sb.WriteString("2) 调用具体工具前，请先确认该工具的参数要求（以当前请求中的工具定义为准）；不确定时先澄清再调用。\n")
 		sb.WriteString("3) 不要臆造不存在的工具名。\n\n")
 	}
+	if s := strings.TrimSpace(injectShellToolGuidance("", names)); s != "" {
+		sb.WriteString(s)
+		sb.WriteString("\n\n")
+	}
 	if s := strings.TrimSpace(instruction); s != "" {
 		sb.WriteString(s)
 	}

internal/multiagent/eino_transient_retry.go

@@ -3,6 +3,7 @@ package multiagent
 import (
 	"context"
 	"errors"
+	"fmt"
 	"strings"
 	"time"
 
@@ -17,8 +18,9 @@ const (
 	defaultEinoRunRetryMaxBackoff  = 30 * time.Second
 )
 
-// isEinoTransientRunError 判断 ADK 运行期错误是否适合指数退避续跑（429、5xx、网络抖动等）。
+// isEinoTransientRunError 是 Eino 运行期「可退避重试 vs 直接失败」的唯一判据。
-// 用户取消、超时、迭代上限等由 run loop 单独处理，不在此列。
+// 429/5xx/网络抖动等返回 true；用户取消、超时、迭代上限、鉴权失败等返回 false。
+// 其它模块（run loop、summarization 等）只调用本函数，不在别处维护平行规则。
 func isEinoTransientRunError(err error) bool {
 	if err == nil {
 		return false
@@ -60,6 +62,7 @@ func isEinoTransientRunError(err error) bool {
 		"dial tcp",
 		"tls handshake timeout",
 		"stream error",
+		"goaway", // http2: server sent GOAWAY and closed the connection
 		"unexpected eof",
 		`": eof`, // net/http: Post "url": EOF (often wraps io.EOF)
 		"unexpected end of json",
@@ -78,6 +81,71 @@ func isEinoTransientRunError(err error) bool {
 	return false
 }
 
+type einoTransientRunRetryPolicy struct {
+	maxAttempts int
+	maxBackoff  time.Duration
+}
+
+func einoTransientRunRetryPolicyFromArgs(args *einoADKRunLoopArgs) einoTransientRunRetryPolicy {
+	return einoTransientRunRetryPolicy{
+		maxAttempts: einoRunRetryMaxAttempts(args),
+		maxBackoff:  einoRunRetryMaxBackoff(args),
+	}
+}
+
+func einoTransientRunRetryPolicyFromMW(mw *config.MultiAgentEinoMiddlewareConfig) einoTransientRunRetryPolicy {
+	maxBackoff := defaultEinoRunRetryMaxBackoff
+	if mw != nil && mw.RunRetryMaxBackoffSec > 0 {
+		maxBackoff = time.Duration(mw.RunRetryMaxBackoffSec) * time.Second
+	}
+	return einoTransientRunRetryPolicy{
+		maxAttempts: RunRetryMaxAttemptsFromConfig(mw),
+		maxBackoff:  maxBackoff,
+	}
+}
+
+// einoTransientRunRetrier 在 run loop 内对临时错误做指数退避并重启 Runner（唯一重试执行层）。
+type einoTransientRunRetrier struct {
+	policy   einoTransientRunRetryPolicy
+	attempts int
+}
+
+func newEinoTransientRunRetrier(policy einoTransientRunRetryPolicy) *einoTransientRunRetrier {
+	return &einoTransientRunRetrier{policy: policy}
+}
+
+// tryRetry 对临时错误退避后返回重启消息；次数用尽返回 exhausted 错误。
+func (r *einoTransientRunRetrier) tryRetry(
+	ctx context.Context,
+	runErr error,
+	args *einoADKRunLoopArgs,
+	baseMsgs, accumulated []adk.Message,
+	baseCount int,
+) (restarted bool, restartMsgs []adk.Message, ctxSource einoRunRestartContextSource, backoff time.Duration, fatal error) {
+	if runErr == nil || !isEinoTransientRunError(runErr) {
+		return false, nil, "", 0, runErr
+	}
+	r.attempts++
+	if r.attempts > r.policy.maxAttempts {
+		return false, nil, "", 0, fmt.Errorf("transient retry exhausted after %d attempts: %w", r.policy.maxAttempts, runErr)
+	}
+	backoff = einoTransientRetryBackoff(r.attempts-1, r.policy.maxBackoff)
+	select {
+	case <-ctx.Done():
+		return false, nil, "", 0, ctx.Err()
+	case <-time.After(backoff):
+	}
+	restartMsgs, ctxSource = einoMessagesForRunRestart(args, baseMsgs, accumulated, baseCount)
+	return true, restartMsgs, ctxSource, backoff, nil
+}
+
+func (r *einoTransientRunRetrier) attempt() int { return r.attempts }
+
+func (r *einoTransientRunRetrier) maxAttempts() int { return r.policy.maxAttempts }
+
+// reset 在退避重试后成功推进（流/消息完整接收）时清零计数，使后续临时错误从第 1 次退避重新开始。
+func (r *einoTransientRunRetrier) reset() { r.attempts = 0 }
+
 func einoRunRetryMaxAttempts(args *einoADKRunLoopArgs) int {
 	if args != nil && args.RunRetryMaxAttempts > 0 {
 		return args.RunRetryMaxAttempts
@@ -85,7 +153,7 @@ func einoRunRetryMaxAttempts(args *einoADKRunLoopArgs) int {
 	return defaultEinoRunRetryMaxAttempts
 }
 
-// RunRetryMaxAttemptsFromConfig 供 handler 分段续跑计数（与 eino_middleware.run_retry_max_attempts 一致）。
+// RunRetryMaxAttemptsFromConfig 与 eino_middleware.run_retry_max_attempts 一致。
 func RunRetryMaxAttemptsFromConfig(mw *config.MultiAgentEinoMiddlewareConfig) int {
 	if mw != nil && mw.RunRetryMaxAttempts > 0 {
 		return mw.RunRetryMaxAttempts
@@ -93,15 +161,6 @@ func RunRetryMaxAttemptsFromConfig(mw *config.MultiAgentEinoMiddlewareConfig) in
 	return defaultEinoRunRetryMaxAttempts
 }
 
-// TransientRetryBackoff 供 handler 在分段续跑前退避。
-func TransientRetryBackoff(attempt int, maxBackoffSec int) time.Duration {
-	max := defaultEinoRunRetryMaxBackoff
-	if maxBackoffSec > 0 {
-		max = time.Duration(maxBackoffSec) * time.Second
-	}
-	return einoTransientRetryBackoff(attempt, max)
-}
-
 func einoRunRetryMaxBackoff(args *einoADKRunLoopArgs) time.Duration {
 	if args != nil && args.RunRetryMaxBackoffSec > 0 {
 		return time.Duration(args.RunRetryMaxBackoffSec) * time.Second
@@ -122,10 +181,11 @@ const (
 // 1) ModelFacingTrace（与模型实际入参一致） 2) 事件流累积的 runAccumulatedMsgs 3) 初始 msgs。
 func einoMessagesForRunRestart(args *einoADKRunLoopArgs, baseMsgs, accumulated []adk.Message, baseCount int) ([]adk.Message, einoRunRestartContextSource) {
 	if trace := persistTraceSource(args, nil); len(trace) > 0 {
-		return append([]adk.Message(nil), trace...), einoRestartContextModelTrace
+		// modelFacingTrace includes prior Instruction system message(s); genModelInput will prepend again.
+		return stripADKSystemMessages(trace), einoRestartContextModelTrace
 	}
 	if len(accumulated) > baseCount {
-		return append([]adk.Message(nil), accumulated...), einoRestartContextAccumulated
+		return stripADKSystemMessages(accumulated), einoRestartContextAccumulated
 	}
 	return append([]adk.Message(nil), baseMsgs...), einoRestartContextInitial
 }

internal/multiagent/eino_transient_retry_test.go

@@ -27,6 +27,7 @@ func TestIsEinoTransientRunError(t *testing.T) {
 		{"429", errors.New("HTTP 429 Too Many Requests"), true},
 		{"rate limit", errors.New(`{"error":"rate limit exceeded"}`), true},
 		{"connection reset", errors.New("read tcp: connection reset by peer"), true},
+		{"http2 goaway", errors.New("failed to receive stream chunk: error, http2: server sent GOAWAY and closed the connection; LastStreamID=791, ErrCode=NO_ERROR"), true},
 		{"unexpected eof", errors.New("unexpected EOF"), true},
 		{"503", errors.New("upstream returned 503"), true},
 		{"iteration limit", errors.New("max iteration reached"), false},
@@ -90,6 +91,46 @@ func TestEinoRunRetryMaxAttemptsFromArgs(t *testing.T) {
 	}
 }
 
+func TestEinoTransientRunRetrierReset(t *testing.T) {
+	t.Parallel()
+	r := newEinoTransientRunRetrier(einoTransientRunRetryPolicy{maxAttempts: 10, maxBackoff: 30 * time.Second})
+	r.attempts = 3
+	r.reset()
+	if r.attempt() != 0 {
+		t.Fatalf("after reset: attempt=%d, want 0", r.attempt())
+	}
+	// 重置后下一次退避应从 2s 起算（attempt index 0）。
+	if got := einoTransientRetryBackoff(r.attempt(), r.policy.maxBackoff); got != 2*time.Second {
+		t.Fatalf("backoff after reset: got %v, want 2s", got)
+	}
+}
+
+func TestEinoTransientRunRetrierConsecutiveFailures(t *testing.T) {
+	t.Parallel()
+	r := newEinoTransientRunRetrier(einoTransientRunRetryPolicy{maxAttempts: 10, maxBackoff: 30 * time.Second})
+	ctx := context.Background()
+	runErr := errors.New("internal server error")
+	args := &einoADKRunLoopArgs{}
+	base := []adk.Message{schema.UserMessage("hi")}
+
+	for want := 1; want <= 3; want++ {
+		restarted, _, _, _, err := r.tryRetry(ctx, runErr, args, base, nil, len(base))
+		if err != nil {
+			t.Fatalf("tryRetry attempt %d: %v", want, err)
+		}
+		if !restarted {
+			t.Fatalf("tryRetry attempt %d: want restarted", want)
+		}
+		if got := r.attempt(); got != want {
+			t.Fatalf("after failure %d: attempt=%d, want %d", want, got, want)
+		}
+	}
+	r.reset()
+	if r.attempt() != 0 {
+		t.Fatalf("after successful recovery reset: attempt=%d, want 0", r.attempt())
+	}
+}
+
 func TestAppendUserMessageIfNeeded(t *testing.T) {
 	t.Parallel()
 	msgs := []adk.Message{schema.UserMessage("old task")}
@@ -102,10 +143,3 @@ func TestAppendUserMessageIfNeeded(t *testing.T) {
 		t.Fatalf("should not duplicate user message: len=%d", len(dup))
 	}
 }
-
-func TestErrTransientRetryContinue(t *testing.T) {
-	t.Parallel()
-	if !errors.Is(ErrTransientRetryContinue, ErrTransientRetryContinue) {
-		t.Fatal("sentinel should match")
-	}
-}

internal/multiagent/execute_exit_error.go

@@ -0,0 +1,15 @@
+package multiagent
+
+import "fmt"
+
+// ExecuteExitError 表示 execute 命令非零退出（预期失败，非超时/中断/流异常）。
+type ExecuteExitError struct {
+	Code int
+}
+
+func (e *ExecuteExitError) Error() string {
+	if e == nil {
+		return "exit status unknown"
+	}
+	return fmt.Sprintf("exit status %d", e.Code)
+}

internal/multiagent/interrupt.go

@@ -5,11 +5,3 @@ import "errors"
 // ErrInterruptContinue 作为 context.CancelCause 使用：用户选择「中断并继续」且当前无进行中的 MCP 工具时，
 // 取消当前推理/流式输出，并在同一会话任务内携带用户补充说明自动续跑下一轮（类似 Hermes 式人机回合）。
 var ErrInterruptContinue = errors.New("agent interrupt: continue with user-supplied context")
-
-// ErrTransientRetryContinue 表示 Run 因 429/网络等临时错误结束，应由 handler 落库轨迹后
-// loadHistoryFromAgentTrace 再开下一轮 Run（与 ErrInterruptContinue 同级的「分段续跑」语义）。
-var ErrTransientRetryContinue = errors.New("agent transient: retry after persisting trace")
-
-// ErrEmptyResponseContinue 表示 Eino ADK 会话正常结束但未捕获到助手正文，应由 handler 落库轨迹后
-// loadHistoryFromAgentTrace 再开下一轮 Run（与 ErrInterruptContinue / ErrTransientRetryContinue 同级）。
-var ErrEmptyResponseContinue = errors.New("agent empty response: continue after persisting trace")

internal/multiagent/orchestrator_instruction.go

@@ -6,6 +6,7 @@ import (
 	"cyberstrike-ai/internal/agents"
 	"cyberstrike-ai/internal/config"
 	"cyberstrike-ai/internal/project"
+	"cyberstrike-ai/internal/projectprompt"
 )
 
 // DefaultPlanExecuteOrchestratorInstruction 当未配置 plan_execute 专用 Markdown / YAML 时的内置主代理（规划/重规划侧）提示。
@@ -122,7 +123,9 @@ func DefaultPlanExecuteOrchestratorInstruction() string {
 
 ## 表达
 
-在调用工具或给出计划变更前，用 2～5 句中文说明当前决策依据与期望证据形态；最终对用户交付结构化结论（发现摘要、证据、风险、下一步）。`
+在调用工具或给出计划变更前，用 2～5 句中文说明当前决策依据与期望证据形态；最终对用户交付结构化结论（发现摘要、证据、风险、下一步）。
+
+` + projectprompt.ShellExecExecuteGuidanceSection()
 }
 
 // DefaultSupervisorOrchestratorInstruction 当未配置 supervisor 专用 Markdown / YAML 时的内置监督者提示（transfer / exit 说明仍由运行时在末尾追加）。

internal/multiagent/orphan_tool_pruner_middleware.go

@@ -27,7 +27,7 @@ import (
 //     本中间件与之互补，专职兜底正向孤儿。
 //   - 仅剔除消息，不向历史里注入虚构 assistant(tc)：虚构 tool_calls 反而会误导模型后续推理。
 //     摘要已覆盖被裁剪段的语义，丢一条原始 tool 结果对对话连贯性影响最小。
-//   - 位置建议：挂在所有可能改写历史的中间件（summarization / reduction / skill / plantask /
+//   - 位置建议：挂在 summarization / reduction / skill / plantask / system 合并 / 续聊 dedup 之后，
 //     tool_search）之后，靠近 ChatModel 调用的那一端。
 type orphanToolPrunerMiddleware struct {
 	adk.BaseChatModelAgentMiddleware

internal/multiagent/runner.go

@@ -20,6 +20,7 @@ import (
 	"cyberstrike-ai/internal/openai"
 	"cyberstrike-ai/internal/project"
 	"cyberstrike-ai/internal/reasoning"
+	"cyberstrike-ai/internal/security"
 
 	einoopenai "github.com/cloudwego/eino-ext/components/model/openai"
 	"github.com/cloudwego/eino/adk"
@@ -120,7 +121,7 @@ func RunDeepAgent(
 		mcpIDs = append(mcpIDs, id)
 		mcpIDsMu.Unlock()
 	}
-	einoExecMonitor := newEinoExecuteMonitorCallback(ag, recorder)
+	einoExecBegin, einoExecFinish := newEinoExecuteMonitorCallbacks(ag, recorder)
 
 	// 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 {
@@ -223,7 +224,7 @@ func RunDeepAgent(
 			}
 			if einoSkillMW != nil {
 				if einoFSTools && einoLoc != nil {
-					subFs, fsErr := subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, id, einoExecMonitor, agentToolTimeoutMinutes(appCfg), nil)
+					subFs, fsErr := subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, id, einoExecBegin, einoExecFinish, agentToolTimeoutMinutes(appCfg), agentShellNoOutputTimeoutSeconds(appCfg), nil)
 					if fsErr != nil {
 						return nil, fmt.Errorf("子代理 %q filesystem 中间件: %w", id, fsErr)
 					}
@@ -231,13 +232,13 @@ func RunDeepAgent(
 				}
 				subHandlers = append(subHandlers, einoSkillMW)
 			}
-			subHandlers = append(subHandlers, subSumMw)
+			subHandlers = appendEinoChatModelTailMiddlewares(subHandlers, einoChatModelTailConfig{
-			// 孤儿 tool 消息兜底：放在 summarization 之后，telemetry 之前，
+				logger:         logger,
-			// 以便 telemetry 记录的 token 数与 LLM 实际入参一致。
+				phase:          "sub_agent:" + id,
-			subHandlers = append(subHandlers, newOrphanToolPrunerMiddleware(logger, "sub_agent:"+id))
+				summarization:  subSumMw,
-			if teleMw := newEinoModelInputTelemetryMiddleware(logger, appCfg.OpenAI.Model, conversationID, "sub_agent"); teleMw != nil {
+				modelName:      appCfg.OpenAI.Model,
-				subHandlers = append(subHandlers, teleMw)
+				conversationID: conversationID,
-			}
+			})
 
 			subInstrFinal := project.AppendVisionImageAnalysisIfReady(instr, appCfg.Vision.Ready())
 			subInstrFinal = injectToolNamesOnlyInstruction(ctx, subInstrFinal, subTools, subToolSearchActive)
@@ -358,12 +359,14 @@ func RunDeepAgent(
 	if einoLoc != nil && einoFSTools {
 		deepBackend = einoLoc
 		deepShell = &einoStreamingShellWrap{
-			inner:              einoLoc,
+			inner:                   security.NewEinoStreamingShell(),
 			invokeNotify:            toolInvokeNotify,
 			einoAgentName:           orchestratorName,
 			outputChunk:             nil,
-			recordMonitor:      einoExecMonitor,
+			beginMonitor:            einoExecBegin,
+			finishMonitor:           einoExecFinish,
 			toolTimeoutMinutes:      agentToolTimeoutMinutes(appCfg),
+			shellNoOutputTimeoutSec: agentShellNoOutputTimeoutSeconds(appCfg),
 		}
 	}
 
@@ -379,14 +382,14 @@ func RunDeepAgent(
 	if einoSkillMW != nil {
 		deepHandlers = append(deepHandlers, einoSkillMW)
 	}
-	deepHandlers = append(deepHandlers, mainSumMw)
+	deepHandlers = appendEinoChatModelTailMiddlewares(deepHandlers, einoChatModelTailConfig{
-	deepHandlers = append(deepHandlers, newOrphanToolPrunerMiddleware(logger, "deep_orchestrator"))
+		logger:         logger,
-	if teleMw := newEinoModelInputTelemetryMiddleware(logger, appCfg.OpenAI.Model, conversationID, "deep_orchestrator"); teleMw != nil {
+		phase:          "deep_orchestrator",
-		deepHandlers = append(deepHandlers, teleMw)
+		summarization:  mainSumMw,
-	}
+		modelName:      appCfg.OpenAI.Model,
-	if capMw := newModelFacingTraceMiddleware(modelFacingTrace); capMw != nil {
+		conversationID: conversationID,
-		deepHandlers = append(deepHandlers, capMw)
+		trace:          modelFacingTrace,
-	}
+	})
 
 	supHandlers := []adk.ChatModelAgentMiddleware{}
 	if len(mainOrchestratorPre) > 0 {
@@ -395,14 +398,14 @@ func RunDeepAgent(
 	if einoSkillMW != nil {
 		supHandlers = append(supHandlers, einoSkillMW)
 	}
-	supHandlers = append(supHandlers, mainSumMw)
+	supHandlers = appendEinoChatModelTailMiddlewares(supHandlers, einoChatModelTailConfig{
-	supHandlers = append(supHandlers, newOrphanToolPrunerMiddleware(logger, "supervisor_orchestrator"))
+		logger:         logger,
-	if teleMw := newEinoModelInputTelemetryMiddleware(logger, appCfg.OpenAI.Model, conversationID, "supervisor_orchestrator"); teleMw != nil {
+		phase:          "supervisor_orchestrator",
-		supHandlers = append(supHandlers, teleMw)
+		summarization:  mainSumMw,
-	}
+		modelName:      appCfg.OpenAI.Model,
-	if capMw := newModelFacingTraceMiddleware(modelFacingTrace); capMw != nil {
+		conversationID: conversationID,
-		supHandlers = append(supHandlers, capMw)
+		trace:          modelFacingTrace,
-	}
+	})
 
 	mainToolsCfg := adk.ToolsConfig{
 		ToolsNodeConfig: compose.ToolsNodeConfig{
@@ -416,7 +419,7 @@ func RunDeepAgent(
 		EmitInternalEvents: true,
 	}
 
-	deepOutKey, modelRetry, taskGen := deepExtrasFromConfig(ma)
+	deepOutKey, taskGen := deepExtrasFromConfig(ma)
 
 	var da adk.Agent
 	switch orchMode {
@@ -428,7 +431,7 @@ func RunDeepAgent(
 		// 构建 filesystem 中间件（与 Deep sub-agent 一致）
 		var peFsMw adk.ChatModelAgentMiddleware
 		if einoSkillMW != nil && einoFSTools && einoLoc != nil {
-			peFsMw, err = subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, "executor", einoExecMonitor, agentToolTimeoutMinutes(appCfg), nil)
+			peFsMw, err = subAgentFilesystemMiddleware(ctx, einoLoc, toolInvokeNotify, "executor", einoExecBegin, einoExecFinish, agentToolTimeoutMinutes(appCfg), agentShellNoOutputTimeoutSeconds(appCfg), nil)
 			if err != nil {
 				return nil, fmt.Errorf("plan_execute filesystem 中间件: %w", err)
 			}
@@ -451,12 +454,14 @@ func RunDeepAgent(
 			SkillMiddleware:      einoSkillMW,
 			FilesystemMiddleware: peFsMw,
 			ModelFacingTrace:     modelFacingTrace,
-			PlannerReplannerRewriteHandlers: []adk.ChatModelAgentMiddleware{
+			PlannerReplannerRewriteHandlers: appendEinoChatModelTailMiddlewares(nil, einoChatModelTailConfig{
-				mainSumMw,
+				logger:         logger,
-				// 孤儿 tool 消息兜底：必须挂在 summarization 之后、telemetry 之前。
+				phase:          "plan_execute_planner_replanner",
-				newOrphanToolPrunerMiddleware(logger, "plan_execute_planner_replanner"),
+				summarization:  mainSumMw,
-				newEinoModelInputTelemetryMiddleware(logger, appCfg.OpenAI.Model, conversationID, "plan_execute_planner_replanner_rewrite"),
+				modelName:      appCfg.OpenAI.Model,
-			},
+				conversationID: conversationID,
+				skipTrace:      true,
+			}),
 		})
 		if perr != nil {
 			return nil, perr
@@ -473,9 +478,6 @@ func RunDeepAgent(
 			Handlers:      supHandlers,
 			Exit:          &adk.ExitTool{},
 		}
-		if modelRetry != nil {
-			supCfg.ModelRetryConfig = modelRetry
-		}
 		if deepOutKey != "" {
 			supCfg.OutputKey = deepOutKey
 		}
@@ -509,9 +511,6 @@ func RunDeepAgent(
 		if deepOutKey != "" {
 			dcfg.OutputKey = deepOutKey
 		}
-		if modelRetry != nil {
-			dcfg.ModelRetryConfig = modelRetry
-		}
 		if taskGen != nil {
 			dcfg.TaskToolDescriptionGenerator = taskGen
 		}

internal/multiagent/shell_tool_guidance.go

@@ -0,0 +1,33 @@
+package multiagent
+
+import (
+	"strings"
+
+	"cyberstrike-ai/internal/projectprompt"
+)
+
+func shellToolsPresent(toolNames []string) bool {
+	for _, n := range toolNames {
+		switch strings.ToLower(strings.TrimSpace(n)) {
+		case "exec", "execute":
+			return true
+		}
+	}
+	return false
+}
+
+// injectShellToolGuidance 在系统提示末尾追加 exec/execute 分工（仅当工具列表含 exec 或 execute）。
+func injectShellToolGuidance(instruction string, toolNames []string) string {
+	if !shellToolsPresent(toolNames) {
+		return instruction
+	}
+	block := strings.TrimSpace(projectprompt.ShellExecExecuteGuidanceSection())
+	if block == "" {
+		return instruction
+	}
+	s := strings.TrimSpace(instruction)
+	if s == "" {
+		return block
+	}
+	return s + "\n\n" + block
+}

internal/multiagent/shell_tool_guidance_test.go

@@ -0,0 +1,17 @@
+package multiagent
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestInjectShellToolGuidance(t *testing.T) {
+	got := injectShellToolGuidance("base", []string{"nmap"})
+	if got != "base" {
+		t.Fatalf("expected unchanged, got %q", got)
+	}
+	got = injectShellToolGuidance("base", []string{"exec", "nmap"})
+	if !strings.Contains(got, "exec/execute") || !strings.Contains(got, "base") {
+		t.Fatalf("expected shell guidance appended, got %q", got)
+	}
+}

internal/multiagent/system_message_normalizer_middleware.go

@@ -0,0 +1,86 @@
+package multiagent
+
+import (
+	"context"
+
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/schema"
+	"go.uber.org/zap"
+)
+
+// systemMessageNormalizerMiddleware merges duplicate role=system messages into a single
+// leading system message before summarization and each ChatModel call.
+type systemMessageNormalizerMiddleware struct {
+	adk.BaseChatModelAgentMiddleware
+	logger *zap.Logger
+	phase  string
+}
+
+func newSystemMessageNormalizerMiddleware(logger *zap.Logger, phase string) adk.ChatModelAgentMiddleware {
+	return &systemMessageNormalizerMiddleware{logger: logger, phase: phase}
+}
+
+func (m *systemMessageNormalizerMiddleware) BeforeModelRewriteState(
+	ctx context.Context,
+	state *adk.ChatModelAgentState,
+	mc *adk.ModelContext,
+) (context.Context, *adk.ChatModelAgentState, error) {
+	_ = mc
+	if m == nil || state == nil || len(state.Messages) == 0 {
+		return ctx, state, nil
+	}
+	before := countADKSystemMessages(state.Messages)
+	if before <= 1 {
+		return ctx, state, nil
+	}
+	normalized := normalizeSingleLeadingSystemMessage(state.Messages, "")
+	if len(normalized) == len(state.Messages) && countADKSystemMessages(normalized) >= before {
+		return ctx, state, nil
+	}
+	if m.logger != nil {
+		m.logger.Info("eino system messages merged",
+			zap.String("phase", m.phase),
+			zap.Int("system_before", before),
+			zap.Int("system_after", countADKSystemMessages(normalized)),
+			zap.Int("messages_before", len(state.Messages)),
+			zap.Int("messages_after", len(normalized)),
+		)
+	}
+	out := *state
+	out.Messages = normalized
+	return ctx, &out, nil
+}
+
+func countADKSystemMessages(msgs []adk.Message) int {
+	n := 0
+	for _, msg := range msgs {
+		if msg != nil && msg.Role == schema.System {
+			n++
+		}
+	}
+	return n
+}
+
+// stripADKSystemMessages removes all system messages. Use before runner.Run restart when
+// genModelInput will prepend a fresh Instruction.
+func stripADKSystemMessages(msgs []adk.Message) []adk.Message {
+	if len(msgs) == 0 {
+		return msgs
+	}
+	out := make([]adk.Message, 0, len(msgs))
+	for _, msg := range msgs {
+		if msg == nil || msg.Role == schema.System {
+			continue
+		}
+		out = append(out, msg)
+	}
+	return out
+}
+
+// mergeCollectedSystemMessages collapses multiple system messages into one (or none).
+func mergeCollectedSystemMessages(systemMsgs []adk.Message) []adk.Message {
+	if len(systemMsgs) == 0 {
+		return nil
+	}
+	return normalizeSingleLeadingSystemMessage(systemMsgs, "")
+}

internal/multiagent/system_message_normalizer_middleware_test.go

@@ -0,0 +1,75 @@
+package multiagent
+
+import (
+	"context"
+	"testing"
+
+	"github.com/cloudwego/eino/adk"
+	"github.com/cloudwego/eino/schema"
+)
+
+func TestStripADKSystemMessages(t *testing.T) {
+	in := []adk.Message{
+		schema.SystemMessage("a"),
+		schema.UserMessage("u"),
+		schema.SystemMessage("b"),
+		schema.AssistantMessage("x", nil),
+	}
+	out := stripADKSystemMessages(in)
+	if len(out) != 2 {
+		t.Fatalf("got %d messages, want 2", len(out))
+	}
+	if out[0].Role != schema.User || out[1].Role != schema.Assistant {
+		t.Fatalf("unexpected roles: %s, %s", out[0].Role, out[1].Role)
+	}
+}
+
+func TestEinoMessagesForRunRestart_StripsSystemFromTrace(t *testing.T) {
+	holder := newModelFacingTraceHolder()
+	holder.storeFromState(&adk.ChatModelAgentState{Messages: []adk.Message{
+		schema.SystemMessage("sys-1"),
+		schema.SystemMessage("sys-2"),
+		schema.UserMessage("task"),
+	}})
+	msgs, src := einoMessagesForRunRestart(&einoADKRunLoopArgs{ModelFacingTrace: holder}, nil, nil, 0)
+	if src != einoRestartContextModelTrace {
+		t.Fatalf("source: got %q want model_trace", src)
+	}
+	if len(msgs) != 1 || msgs[0].Role != schema.User {
+		t.Fatalf("expected user-only restart msgs, got %+v", msgs)
+	}
+}
+
+func TestSystemMessageNormalizerMiddleware_MergesDuplicates(t *testing.T) {
+	mw := newSystemMessageNormalizerMiddleware(nil, "test")
+	state := &adk.ChatModelAgentState{Messages: []adk.Message{
+		schema.SystemMessage("a"),
+		schema.SystemMessage("b"),
+		schema.UserMessage("u"),
+	}}
+	_, out, err := mw.(*systemMessageNormalizerMiddleware).BeforeModelRewriteState(context.Background(), state, nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if countADKSystemMessages(out.Messages) != 1 {
+		t.Fatalf("want 1 system, got %d", countADKSystemMessages(out.Messages))
+	}
+	if out.Messages[0].Content != "a\n\nb" {
+		t.Fatalf("merged content: %q", out.Messages[0].Content)
+	}
+}
+
+func TestSystemMessageNormalizerMiddleware_NoOpSingleSystem(t *testing.T) {
+	mw := newSystemMessageNormalizerMiddleware(nil, "test")
+	state := &adk.ChatModelAgentState{Messages: []adk.Message{
+		schema.SystemMessage("only"),
+		schema.UserMessage("u"),
+	}}
+	_, out, err := mw.(*systemMessageNormalizerMiddleware).BeforeModelRewriteState(context.Background(), state, nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if out != state {
+		t.Fatalf("expected same state pointer for no-op")
+	}
+}

internal/project/workspace.go

@@ -0,0 +1,68 @@
+package project
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"strings"
+)
+
+func sanitizeWorkspacePathSegment(s string) string {
+	s = strings.TrimSpace(s)
+	if s == "" {
+		return "default"
+	}
+	s = strings.ReplaceAll(s, string(filepath.Separator), "-")
+	s = strings.ReplaceAll(s, "/", "-")
+	s = strings.ReplaceAll(s, "\\", "-")
+	s = strings.ReplaceAll(s, "..", "__")
+	if len(s) > 180 {
+		s = s[:180]
+	}
+	return s
+}
+
+// WorkspaceRootDir returns the relative workspace root for downloads and local analysis.
+// Project-bound sessions share projects/<id>/; otherwise conversations/<id>/.
+func WorkspaceRootDir(configuredBase, projectID, conversationID string) string {
+	base := strings.TrimSpace(configuredBase)
+	if base == "" {
+		base = filepath.Join("tmp", "workspace")
+	}
+	if pid := strings.TrimSpace(projectID); pid != "" {
+		return filepath.Join(base, "projects", sanitizeWorkspacePathSegment(pid))
+	}
+	conv := strings.TrimSpace(conversationID)
+	if conv == "" {
+		conv = "default"
+	}
+	return filepath.Join(base, "conversations", sanitizeWorkspacePathSegment(conv))
+}
+
+// EnsureWorkspace creates the workspace directory and returns its absolute path.
+func EnsureWorkspace(root string) (string, error) {
+	abs, err := filepath.Abs(strings.TrimSpace(root))
+	if err != nil {
+		return "", fmt.Errorf("workspace abs: %w", err)
+	}
+	if err := os.MkdirAll(abs, 0o755); err != nil {
+		return "", fmt.Errorf("workspace mkdir: %w", err)
+	}
+	return abs, nil
+}
+
+// BuildWorkspaceBlock instructs the agent to use the session workspace instead of /tmp.
+func BuildWorkspaceBlock(absPath string) string {
+	absPath = strings.TrimSpace(absPath)
+	if absPath == "" {
+		return ""
+	}
+	return fmt.Sprintf(`## 会话工作目录（下载与本地分析）
+
+**必须使用以下目录**保存 curl/wget 下载的文件、临时 HTML/JS，以及 read_file/glob/grep 的检索范围：
+`+"`%s`"+`
+
+- **禁止**使用系统 `+"`/tmp`"+` 或其它全局临时目录（多项目/多会话会互窜遗留文件）。
+- 下载示例：`+"`curl -o '%s/page.html' 'https://target/'`"+`；exec 时可将 `+"`workdir`"+` 设为该目录。
+- 读取前用 glob/grep/read_file **限定在该目录**下搜索，勿在 `+"`/tmp`"+` 盲目检索。`, absPath, absPath)
+}

internal/project/workspace_test.go

@@ -0,0 +1,52 @@
+package project
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+func TestWorkspaceRootDirProjectScoped(t *testing.T) {
+	got := WorkspaceRootDir("", "proj-1", "conv-1")
+	want := filepath.Join("tmp", "workspace", "projects", "proj-1")
+	if got != want {
+		t.Fatalf("got %q want %q", got, want)
+	}
+}
+
+func TestWorkspaceRootDirConversationScoped(t *testing.T) {
+	got := WorkspaceRootDir("/data/ws", "", "conv-abc")
+	want := filepath.Join("/data/ws", "conversations", "conv-abc")
+	if got != want {
+		t.Fatalf("got %q want %q", got, want)
+	}
+}
+
+func TestEnsureWorkspaceCreatesDir(t *testing.T) {
+	root := filepath.Join(t.TempDir(), "nested", "workspace")
+	abs, err := EnsureWorkspace(root)
+	if err != nil {
+		t.Fatalf("EnsureWorkspace: %v", err)
+	}
+	st, err := os.Stat(abs)
+	if err != nil {
+		t.Fatalf("Stat: %v", err)
+	}
+	if !st.IsDir() {
+		t.Fatal("expected directory")
+	}
+}
+
+func TestBuildWorkspaceBlockMentionsPath(t *testing.T) {
+	block := BuildWorkspaceBlock("/opt/csai/tmp/workspace/projects/p1")
+	if block == "" {
+		t.Fatal("expected non-empty block")
+	}
+	if !strings.Contains(block, "/opt/csai/tmp/workspace/projects/p1") {
+		t.Fatalf("block missing path: %s", block)
+	}
+	if !strings.Contains(block, "/tmp") {
+		t.Fatalf("block should warn about /tmp: %s", block)
+	}
+}

internal/projectprompt/shell_tools.go

@@ -0,0 +1,11 @@
+package projectprompt
+
+// ShellExecExecuteGuidanceSection 供单代理/多代理系统提示追加：exec 与 execute 分工（尽量短）。
+func ShellExecExecuteGuidanceSection() string {
+	return `Shell（exec/execute）：有专用 MCP 工具时优先专用工具；系统命令（管道、workdir、后台 &）用 exec；skills/ 内脚本（配合 read_file、skill）用 execute；多步扫描分拆调用，禁止一条 shell 串多个扫描器。下载/临时文件须写入系统提示中的「会话工作目录」，禁止用 /tmp。`
+}
+
+// ShellExecExecuteGuidanceReconSuffix 侦察子代理可选追加（一行）。
+func ShellExecExecuteGuidanceReconSuffix() string {
+	return `枚举优先 subfinder、amass 等专用 MCP，勿 exec/execute 拼长链。`
+}

internal/security/command_failure_format.go

@@ -0,0 +1,56 @@
+package security
+
+import (
+	"errors"
+	"fmt"
+	"os/exec"
+	"strings"
+)
+
+// FormatCommandFailureResult 与 exec 工具 ToolResult 文案一致（不含 ToolErrorPrefix）。
+func FormatCommandFailureResult(exitCode int, output string) string {
+	output = strings.TrimSpace(output)
+	errMsg := fmt.Sprintf("exit status %d", exitCode)
+	if output == "" {
+		return fmt.Sprintf("命令执行失败: %s", errMsg)
+	}
+	if strings.HasPrefix(output, "命令执行失败:") {
+		return output
+	}
+	return fmt.Sprintf("命令执行失败: %s\n输出: %s", errMsg, output)
+}
+
+// FormatCommandFailureFromErr 根据 exec/execute 返回的 error 生成统一失败文案（IsError 正文）。
+func FormatCommandFailureFromErr(err error, output string) string {
+	if err == nil {
+		return strings.TrimSpace(output)
+	}
+	var exitError *exec.ExitError
+	if errors.As(err, &exitError) {
+		return FormatCommandFailureResult(exitError.ExitCode(), output)
+	}
+	output = strings.TrimSpace(output)
+	if output == "" {
+		return fmt.Sprintf("命令执行失败: %v", err)
+	}
+	if strings.HasPrefix(output, "命令执行失败:") {
+		return output
+	}
+	return fmt.Sprintf("命令执行失败: %v\n输出: %s", err, output)
+}
+
+// ExecuteFailureStatusLine 流式 execute 结束时追加的单行状态（输出正文已在流中推送过）。
+func ExecuteFailureStatusLine(exitCode int) string {
+	return fmt.Sprintf("\n命令执行失败: exit status %d", exitCode)
+}
+
+// IsCommandFailureResult 判断工具结果正文是否表示命令非零退出（用于 execute / exec 对齐 isError）。
+func IsCommandFailureResult(content string) bool {
+	return strings.Contains(content, "命令执行失败:")
+}
+
+// IsLegacyShellExitNoise 过滤旧版 shell 流中冗余的 exit code 行。
+func IsLegacyShellExitNoise(s string) bool {
+	trimmed := strings.TrimSpace(s)
+	return strings.HasPrefix(trimmed, "command exited with non-zero code ")
+}

internal/security/command_failure_format_test.go

@@ -0,0 +1,54 @@
+package security
+
+import (
+	"errors"
+	"os/exec"
+	"strings"
+	"testing"
+)
+
+func TestFormatCommandFailureResult(t *testing.T) {
+	got := FormatCommandFailureResult(1, "sudo: password required")
+	want := "命令执行失败: exit status 1\n输出: sudo: password required"
+	if got != want {
+		t.Fatalf("got %q want %q", got, want)
+	}
+	if FormatCommandFailureResult(2, "") != "命令执行失败: exit status 2" {
+		t.Fatal("empty output format")
+	}
+	if FormatCommandFailureResult(1, "命令执行失败: exit status 1") != "命令执行失败: exit status 1" {
+		t.Fatal("should not double-wrap")
+	}
+}
+
+func TestIsCommandFailureResult(t *testing.T) {
+	if !IsCommandFailureResult("sudo: err\n命令执行失败: exit status 1") {
+		t.Fatal("expected true")
+	}
+	if IsCommandFailureResult("sudo: err only") {
+		t.Fatal("expected false")
+	}
+}
+
+func TestFormatCommandFailureFromErr(t *testing.T) {
+	cmd := exec.Command("sh", "-c", "exit 42")
+	err := cmd.Run()
+	got := FormatCommandFailureFromErr(err, "oops")
+	if got != "命令执行失败: exit status 42\n输出: oops" {
+		t.Fatalf("got %q", got)
+	}
+	timeoutErr := errors.New("shell inactivity timeout (300s)")
+	got2 := FormatCommandFailureFromErr(timeoutErr, "already timed out")
+	if !strings.Contains(got2, "shell inactivity timeout") || !strings.Contains(got2, "already timed out") {
+		t.Fatalf("got %q", got2)
+	}
+}
+
+func TestIsLegacyShellExitNoise(t *testing.T) {
+	if !IsLegacyShellExitNoise("command exited with non-zero code 1\n") {
+		t.Fatal("expected legacy noise")
+	}
+	if IsLegacyShellExitNoise("sudo: failed") {
+		t.Fatal("unexpected noise")
+	}
+}

internal/security/executor.go

@@ -36,6 +36,7 @@ type Executor struct {
 	toolIndex               map[string]*config.ToolConfig // 工具索引，用于 O(1) 查找
 	mcpServer               *mcp.Server
 	logger                  *zap.Logger
+	shellNoOutputTimeoutSec int // execute/exec 无新输出空闲秒数；0=默认 300；-1=关闭（见 SetShellNoOutputTimeoutSeconds）
 }
 
 // NewExecutor 创建新的执行器
@@ -51,6 +52,11 @@ func NewExecutor(cfg *config.SecurityConfig, mcpServer *mcp.Server, logger *zap.
 	return executor
 }
 
+// SetShellNoOutputTimeoutSeconds 配置 exec 工具无输出空闲终止（与 agent.shell_no_output_timeout_seconds 一致）。
+func (e *Executor) SetShellNoOutputTimeoutSeconds(sec int) {
+	e.shellNoOutputTimeoutSec = sec
+}
+
 // buildToolIndex 构建工具索引，将 O(n) 查找优化为 O(1)
 func (e *Executor) buildToolIndex() {
 	e.toolIndex = make(map[string]*config.ToolConfig)
@@ -133,6 +139,7 @@ func (e *Executor) ExecuteTool(ctx context.Context, toolName string, args map[st
 	// 执行命令
 	cmd := exec.CommandContext(ctx, toolConfig.Command, cmdArgs...)
 	applyDefaultTerminalEnv(cmd)
+	attachNonInteractiveStdin(cmd)
 	_ = prepareShellCmdSession(cmd)
 
 	e.logger.Info("执行安全工具",
@@ -144,7 +151,7 @@ func (e *Executor) ExecuteTool(ctx context.Context, toolName string, args map[st
 	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(ctx, cmd, cb)
+		output, err = streamCommandOutput(ctx, cmd, cb, ResolveShellNoOutputTimeoutSeconds(e.shellNoOutputTimeoutSec))
 		if err != nil && shouldRetryWithPTY(output) {
 			e.logger.Info("检测到工具需要 TTY，使用 PTY 重试",
 				zap.String("tool", toolName),
@@ -797,6 +804,8 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 		zap.String("command", command),
 	)
 
+	command = PrepareNonInteractiveShellCommand(command)
+
 	// 获取shell类型（可选，默认为sh）
 	shell := "sh"
 	if s, ok := args["shell"].(string); ok && s != "" {
@@ -820,8 +829,7 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 	} else {
 		cmd = exec.CommandContext(ctx, shell, "-c", command)
 	}
-	applyDefaultTerminalEnv(cmd)
+	ConfigureShellCmdForAgentExecute(cmd)
-	_ = prepareShellCmdSession(cmd)
 
 	// 执行命令
 	e.logger.Info("执行系统命令",
@@ -850,8 +858,7 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 		} else {
 			pidCmd = exec.CommandContext(ctx, shell, "-c", pidCommand)
 		}
-		applyDefaultTerminalEnv(pidCmd)
+		ConfigureShellCmdForAgentExecute(pidCmd)
-		_ = prepareShellCmdSession(pidCmd)
 
 		// 获取stdout管道
 		stdout, err := pidCmd.StdoutPipe()
@@ -963,15 +970,14 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 	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(ctx, cmd, cb)
+		output, err = streamCommandOutput(ctx, cmd, cb, ResolveShellNoOutputTimeoutSeconds(e.shellNoOutputTimeoutSec))
 		if err != nil && shouldRetryWithPTY(output) {
 			e.logger.Info("检测到系统命令需要 TTY，使用 PTY 重试")
 			cmd2 := exec.CommandContext(ctx, shell, "-c", command)
 			if workDir != "" {
 				cmd2.Dir = workDir
 			}
-			applyDefaultTerminalEnv(cmd2)
+			ConfigureShellCmdForAgentExecute(cmd2)
-			_ = prepareShellCmdSession(cmd2)
 			output, err = runCommandWithPTY(ctx, cmd2, cb)
 		}
 	} else {
@@ -984,8 +990,7 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 			if workDir != "" {
 				cmd2.Dir = workDir
 			}
-			applyDefaultTerminalEnv(cmd2)
+			ConfigureShellCmdForAgentExecute(cmd2)
-			_ = prepareShellCmdSession(cmd2)
 			output, err = runCommandWithPTY(ctx, cmd2, nil)
 		}
 	}
@@ -999,7 +1004,7 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 			Content: []mcp.Content{
 				{
 					Type: "text",
-					Text: fmt.Sprintf("命令执行失败: %v\n输出: %s", err, string(output)),
+					Text: FormatCommandFailureFromErr(err, output),
 				},
 			},
 			IsError: true,
@@ -1024,7 +1029,7 @@ func (e *Executor) executeSystemCommand(ctx context.Context, args map[string]int
 
 // streamCommandOutput 以“边读边回调”的方式读取命令 stdout/stderr。
 // 使用定长块读取，避免按行读取在无换行输出时永久阻塞；ctx 取消时终止进程树。
-func streamCommandOutput(ctx context.Context, cmd *exec.Cmd, cb ToolOutputCallback) (string, error) {
+func streamCommandOutput(ctx context.Context, cmd *exec.Cmd, cb ToolOutputCallback, noOutputSec int) (string, error) {
 	if err := prepareShellCmdSession(cmd); err != nil {
 		return "", err
 	}
@@ -1091,14 +1096,45 @@ func streamCommandOutput(ctx context.Context, cmd *exec.Cmd, cb ToolOutputCallba
 		lastFlush = time.Now()
 	}
 
-	for chunk := range chunks {
+	idleWatch := NewShellInactivityWatch(noOutputSec)
+	if idleWatch != nil {
+		defer idleWatch.Stop()
+	}
+
+	fireInactivity := func() {
+		terminateCmdTree(cmd)
+		msg := ShellNoOutputTimeoutMessage(idleWatch.Sec)
+		outBuilder.WriteString(msg)
+		if cb != nil {
+			cb(msg)
+		}
+		_ = cmd.Wait()
+	}
+
+chunksLoop:
+	for {
+		var idleCh <-chan struct{}
+		if idleWatch != nil {
+			idleCh = idleWatch.Expired
+		}
+		select {
+		case <-idleCh:
+			fireInactivity()
+			return outBuilder.String(), fmt.Errorf("shell inactivity timeout (%ds)", idleWatch.Sec)
+		case chunk, ok := <-chunks:
+			if !ok {
+				break chunksLoop
+			}
+			if chunk != "" && idleWatch != nil {
+				idleWatch.Bump()
+			}
 			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
@@ -1116,6 +1152,7 @@ func applyDefaultTerminalEnv(cmd *exec.Cmd) {
 	if cmd.Env == nil {
 		cmd.Env = os.Environ()
 	}
+	cmd.Env = ApplyNonInteractivePagerEnv(cmd.Env)
 	// 如果用户已设置 TERM/COLUMNS/LINES，则不覆盖
 	has := func(k string) bool {
 		prefix := k + "="
@@ -1159,7 +1196,7 @@ func runCommandWithPTY(ctx context.Context, cmd *exec.Cmd, cb ToolOutputCallback
 	if runtime.GOOS == "windows" {
 		// PTY 方案为类 Unix；Windows 走原逻辑
 		if cb != nil {
-			return streamCommandOutput(ctx, cmd, cb)
+			return streamCommandOutput(ctx, cmd, cb, 0)
 		}
 		_ = prepareShellCmdSession(cmd)
 		out, err := cmd.CombinedOutput()

internal/security/executor_test.go

@@ -71,6 +71,27 @@ func TestExecuteSystemCommand_BackgroundDoesNotBlockOnChildStdout(t *testing.T)
 	}
 }
 
+func TestExecuteSystemCommand_FailureFormat(t *testing.T) {
+	executor, _ := setupTestExecutor(t)
+	res, err := executor.executeSystemCommand(context.Background(), map[string]interface{}{
+		"command": "echo fail-msg >&2; exit 7",
+		"shell":   "sh",
+	})
+	if err != nil {
+		t.Fatalf("executeSystemCommand: %v", err)
+	}
+	if res == nil || !res.IsError {
+		t.Fatalf("expected IsError, got %+v", res)
+	}
+	text := res.Content[0].Text
+	if text != FormatCommandFailureResult(7, "fail-msg\n") && text != FormatCommandFailureResult(7, "fail-msg") {
+		t.Fatalf("unexpected failure text: %q", text)
+	}
+	if !strings.Contains(text, "exit status 7") || !strings.Contains(text, "fail-msg") {
+		t.Fatalf("unexpected failure text: %q", text)
+	}
+}
+
 func TestBuildCommandArgs_NmapSkipsEmptyOptionalFlags(t *testing.T) {
 	pos1 := 1
 	executor, _ := setupTestExecutor(t)

internal/security/shell_execute_stream.go

@@ -0,0 +1,200 @@
+package security
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"os/exec"
+	"sync"
+
+	"github.com/cloudwego/eino/adk/filesystem"
+	"github.com/cloudwego/eino/schema"
+)
+
+// ConfigureShellCmdForAgentExecute 与 exec 工具一致：非交互 stdin、pager/TERM 环境、独立进程组。
+func ConfigureShellCmdForAgentExecute(cmd *exec.Cmd) {
+	if cmd == nil {
+		return
+	}
+	applyDefaultTerminalEnv(cmd)
+	attachNonInteractiveStdin(cmd)
+	_ = prepareShellCmdSession(cmd)
+}
+
+// TerminateShellCmdTree 尽力终止 shell 及其子进程组（与 exec/execute 超时取消一致）。
+func TerminateShellCmdTree(cmd *exec.Cmd) {
+	terminateCmdTree(cmd)
+}
+
+// EinoStreamingShell 为 Eino ADK execute 工具提供流式 shell，行为与 exec 对齐：
+// 并发读取 stdout/stderr（定长块，非按行），避免官方 local.ExecuteStreaming 先排空 stdout
+// 导致 stderr 错误（如 sudo 密码提示）长时间不可见、UI 一直显示「执行中」。
+type EinoStreamingShell struct{}
+
+// NewEinoStreamingShell 创建 execute 流式 shell 实现。
+func NewEinoStreamingShell() *EinoStreamingShell {
+	return &EinoStreamingShell{}
+}
+
+// ExecuteStreaming 实现 filesystem.StreamingShell。
+func (s *EinoStreamingShell) ExecuteStreaming(ctx context.Context, input *filesystem.ExecuteRequest) (*schema.StreamReader[*filesystem.ExecuteResponse], error) {
+	if input == nil || input.Command == "" {
+		return nil, fmt.Errorf("command is required")
+	}
+
+	sr, w := schema.Pipe[*filesystem.ExecuteResponse](100)
+	if input.RunInBackendGround {
+		go runShellInBackground(ctx, input.Command, w)
+		return sr, nil
+	}
+	go streamShellForeground(ctx, input.Command, w)
+	return sr, nil
+}
+
+func runShellInBackground(ctx context.Context, command string, w *schema.StreamWriter[*filesystem.ExecuteResponse]) {
+	defer w.Close()
+
+	cmd := exec.CommandContext(ctx, "/bin/sh", "-c", command)
+	ConfigureShellCmdForAgentExecute(cmd)
+	stdout, err := cmd.StdoutPipe()
+	if err != nil {
+		_ = w.Send(nil, fmt.Errorf("failed to create stdout pipe: %w", err))
+		return
+	}
+	stderr, err := cmd.StderrPipe()
+	if err != nil {
+		_ = stdout.Close()
+		_ = w.Send(nil, fmt.Errorf("failed to create stderr pipe: %w", err))
+		return
+	}
+	if err := cmd.Start(); err != nil {
+		_ = stdout.Close()
+		_ = stderr.Close()
+		_ = w.Send(nil, fmt.Errorf("failed to start command: %w", err))
+		return
+	}
+
+	done := make(chan struct{})
+	go func() {
+		drainShellPipes(stdout, stderr)
+		_ = cmd.Wait()
+		close(done)
+	}()
+
+	select {
+	case <-done:
+	case <-ctx.Done():
+		TerminateShellCmdTree(cmd)
+	}
+
+	exitCode := 0
+	_ = w.Send(&filesystem.ExecuteResponse{
+		Output:   "command started in background\n",
+		ExitCode: &exitCode,
+	}, nil)
+}
+
+func drainShellPipes(stdout, stderr io.Reader) {
+	var wg sync.WaitGroup
+	wg.Add(2)
+	go func() {
+		defer wg.Done()
+		_, _ = io.Copy(io.Discard, stdout)
+	}()
+	go func() {
+		defer wg.Done()
+		_, _ = io.Copy(io.Discard, stderr)
+	}()
+	wg.Wait()
+}
+
+func streamShellForeground(ctx context.Context, command string, w *schema.StreamWriter[*filesystem.ExecuteResponse]) {
+	defer w.Close()
+
+	cmd := exec.CommandContext(ctx, "/bin/sh", "-c", command)
+	ConfigureShellCmdForAgentExecute(cmd)
+
+	stdoutPipe, err := cmd.StdoutPipe()
+	if err != nil {
+		_ = w.Send(nil, fmt.Errorf("failed to create stdout pipe: %w", err))
+		return
+	}
+	stderrPipe, err := cmd.StderrPipe()
+	if err != nil {
+		_ = stdoutPipe.Close()
+		_ = w.Send(nil, fmt.Errorf("failed to create stderr pipe: %w", err))
+		return
+	}
+	if err := cmd.Start(); err != nil {
+		_ = stdoutPipe.Close()
+		_ = stderrPipe.Close()
+		_ = w.Send(nil, fmt.Errorf("failed to start command: %w", err))
+		return
+	}
+
+	stopWatch := make(chan struct{})
+	go func() {
+		select {
+		case <-ctx.Done():
+			TerminateShellCmdTree(cmd)
+		case <-stopWatch:
+		}
+	}()
+	defer close(stopWatch)
+
+	chunks := make(chan string, 64)
+	var wg sync.WaitGroup
+	readFn := func(r io.Reader) {
+		defer wg.Done()
+		buf := make([]byte, 8192)
+		for {
+			n, readErr := r.Read(buf)
+			if n > 0 {
+				chunks <- string(buf[:n])
+			}
+			if readErr != nil {
+				return
+			}
+		}
+	}
+
+	wg.Add(2)
+	go readFn(stdoutPipe)
+	go readFn(stderrPipe)
+	go func() {
+		wg.Wait()
+		close(chunks)
+	}()
+
+	hadOutput := false
+	for chunk := range chunks {
+		if chunk == "" {
+			continue
+		}
+		hadOutput = true
+		if w.Send(&filesystem.ExecuteResponse{Output: chunk}, nil) {
+			TerminateShellCmdTree(cmd)
+			return
+		}
+	}
+
+	waitErr := cmd.Wait()
+	if waitErr == nil {
+		exitCode := 0
+		_ = w.Send(&filesystem.ExecuteResponse{ExitCode: &exitCode}, nil)
+		return
+	}
+
+	var exitError *exec.ExitError
+	if errors.As(waitErr, &exitError) {
+		exitCode := exitError.ExitCode()
+		resp := &filesystem.ExecuteResponse{ExitCode: &exitCode}
+		if !hadOutput {
+			resp.Output = FormatCommandFailureResult(exitCode, "")
+		}
+		_ = w.Send(resp, nil)
+		return
+	}
+	_ = w.Send(nil, fmt.Errorf("command failed: %w", waitErr))
+}

internal/security/shell_execute_stream_test.go

@@ -0,0 +1,117 @@
+package security
+
+import (
+	"context"
+	"errors"
+	"io"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/cloudwego/eino/adk/filesystem"
+)
+
+func TestEinoStreamingShell_StreamsStderrBeforeStdoutEOF(t *testing.T) {
+	shell := NewEinoStreamingShell()
+	cmd := PrepareNonInteractiveShellCommand("echo err-only >&2; exit 1")
+	sr, err := shell.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: cmd})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	start := time.Now()
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+		if resp != nil && resp.Output != "" {
+			got.WriteString(resp.Output)
+		}
+	}
+	if time.Since(start) > 3*time.Second {
+		t.Fatalf("expected fast completion, took %v", time.Since(start))
+	}
+	if !strings.Contains(got.String(), "err-only") {
+		t.Fatalf("expected stderr in output, got: %q", got.String())
+	}
+}
+
+func TestEinoStreamingShell_SudoFailsFast(t *testing.T) {
+	shell := NewEinoStreamingShell()
+	cmd := PrepareNonInteractiveShellCommand("sudo whoami && sudo cat /etc/os-release")
+	sr, err := shell.ExecuteStreaming(context.Background(), &filesystem.ExecuteRequest{Command: cmd})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	start := time.Now()
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			t.Fatalf("recv: %v", rerr)
+		}
+		if resp == nil {
+			continue
+		}
+		got.WriteString(resp.Output)
+	}
+	if time.Since(start) > 5*time.Second {
+		t.Fatalf("sudo should fail quickly, took %v output=%q", time.Since(start), got.String())
+	}
+	out := got.String()
+	if strings.Contains(out, "command exited with non-zero code") {
+		t.Fatalf("legacy exit line present: %q", out)
+	}
+	if !strings.Contains(out, "sudo") && !strings.Contains(out, "password") && !strings.Contains(out, "terminal") {
+		t.Fatalf("expected sudo error text, got: %q", out)
+	}
+}
+
+func TestEinoStreamingShell_StderrWhileStdoutBlocks(t *testing.T) {
+	shell := NewEinoStreamingShell()
+	// 模拟 sudo：stderr 先有输出，stdout 侧进程仍挂起；旧 eino local 在首包 stderr 前不会向流写任何内容。
+	cmd := PrepareNonInteractiveShellCommand(`echo "password prompt" >&2; sleep 30`)
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Second)
+	defer cancel()
+
+	sr, err := shell.ExecuteStreaming(ctx, &filesystem.ExecuteRequest{Command: cmd})
+	if err != nil {
+		t.Fatalf("ExecuteStreaming: %v", err)
+	}
+	defer sr.Close()
+
+	start := time.Now()
+	var got strings.Builder
+	for {
+		resp, rerr := sr.Recv()
+		if errors.Is(rerr, io.EOF) {
+			break
+		}
+		if rerr != nil {
+			break
+		}
+		if resp != nil && resp.Output != "" {
+			got.WriteString(resp.Output)
+			if strings.Contains(got.String(), "password prompt") {
+				break
+			}
+		}
+	}
+	if time.Since(start) > 1500*time.Millisecond {
+		t.Fatalf("expected stderr promptly, took %v output=%q", time.Since(start), got.String())
+	}
+	if !strings.Contains(got.String(), "password prompt") {
+		t.Fatalf("expected early stderr, got: %q", got.String())
+	}
+}

internal/security/shell_noninteractive.go

@@ -0,0 +1,163 @@
+package security
+
+import (
+	"fmt"
+	"os"
+	"os/exec"
+	"strings"
+	"sync"
+	"time"
+)
+
+// ShellNoOutputTimeoutMessage 长时间无新 stdout/stderr 时的提示（软失败，模型可见）。
+func ShellNoOutputTimeoutMessage(idleSec int) string {
+	return fmt.Sprintf(`命令已终止：超过 %d 秒没有新的输出，疑似在等待交互输入或已挂起。
+
+长时静默任务请使用末尾 & 后台运行，或增大 agent.shell_no_output_timeout_seconds（-1=关闭此检测）。
+
+Command terminated: no new output for %d seconds (possible interactive wait or hung process).`, idleSec, idleSec)
+}
+
+// ShellInactivityWatch 在 noOutputSec 内无任何新输出时向 expired 发送信号；每次 Bump 重置计时。
+// 与「仅有首包输出就永久取消计时」不同，可兜住 sudo 打印 Password 提示后继续挂起等情况。
+type ShellInactivityWatch struct {
+	Sec     int
+	mu      sync.Mutex
+	timer   *time.Timer
+	Expired chan struct{}
+}
+
+func NewShellInactivityWatch(noOutputSec int) *ShellInactivityWatch {
+	sec := ResolveShellNoOutputTimeoutSeconds(noOutputSec)
+	if sec <= 0 {
+		return nil
+	}
+	w := &ShellInactivityWatch{
+		Sec:     sec,
+		Expired: make(chan struct{}, 1),
+	}
+	w.Bump()
+	return w
+}
+
+func (w *ShellInactivityWatch) Bump() {
+	if w == nil || w.Sec <= 0 {
+		return
+	}
+	w.mu.Lock()
+	defer w.mu.Unlock()
+	if w.timer != nil {
+		w.timer.Stop()
+	}
+	w.timer = time.AfterFunc(time.Duration(w.Sec)*time.Second, func() {
+		select {
+		case w.Expired <- struct{}{}:
+		default:
+		}
+	})
+}
+
+func (w *ShellInactivityWatch) Stop() {
+	if w == nil {
+		return
+	}
+	w.mu.Lock()
+	defer w.mu.Unlock()
+	if w.timer != nil {
+		w.timer.Stop()
+		w.timer = nil
+	}
+}
+
+// ResolveShellNoOutputTimeoutSeconds：0=默认 300（5 分钟）；-1=关闭；>0=自定义。
+func ResolveShellNoOutputTimeoutSeconds(sec int) int {
+	if sec < 0 {
+		return 0
+	}
+	if sec == 0 {
+		return 300
+	}
+	return sec
+}
+
+// PrependNonInteractiveShellExports 为 sh -c 注入通用非交互环境（pager 等），不维护命令黑名单。
+func PrependNonInteractiveShellExports(shellCommand string) string {
+	if strings.TrimSpace(shellCommand) == "" {
+		return shellCommand
+	}
+	upper := strings.ToUpper(shellCommand)
+	var pairs []string
+	add := func(key, val string) {
+		if strings.Contains(upper, strings.ToUpper(key)) {
+			return
+		}
+		pairs = append(pairs, key+"="+val)
+	}
+	add("GIT_PAGER", "cat")
+	add("PAGER", "cat")
+	add("SYSTEMD_PAGER", "cat")
+	add("DEBIAN_FRONTEND", "noninteractive")
+	if len(pairs) == 0 {
+		return shellCommand
+	}
+	return "export " + strings.Join(pairs, " ") + "\n" + shellCommand
+}
+
+// PrependNonInteractiveStdinRedirect 为 sh -c 关闭 stdin（与 attachNonInteractiveStdin 等价），
+// 使 read/input()/sudo -S 等从 stdin 读取的程序快速失败而非挂起。已含 </dev/null 时不重复注入。
+func PrependNonInteractiveStdinRedirect(shellCommand string) string {
+	if strings.TrimSpace(shellCommand) == "" {
+		return shellCommand
+	}
+	lower := strings.ToLower(shellCommand)
+	if strings.Contains(lower, "</dev/null") || strings.Contains(lower, "0</dev/null") {
+		return shellCommand
+	}
+	return "exec </dev/null\n" + shellCommand
+}
+
+// PrepareNonInteractiveShellCommand 组合非交互包装：stdin 关闭 + pager 等环境变量（零名单）。
+func PrepareNonInteractiveShellCommand(shellCommand string) string {
+	return PrependNonInteractiveStdinRedirect(PrependNonInteractiveShellExports(shellCommand))
+}
+
+// ApplyNonInteractivePagerEnv 为 exec.Cmd 补齐与 PrependNonInteractiveShellExports 一致的环境变量。
+func ApplyNonInteractivePagerEnv(cmdEnv []string) []string {
+	if cmdEnv == nil {
+		cmdEnv = []string{}
+	}
+	has := func(k string) bool {
+		prefix := k + "="
+		for _, e := range cmdEnv {
+			if strings.HasPrefix(e, prefix) {
+				return true
+			}
+		}
+		return false
+	}
+	if !has("GIT_PAGER") {
+		cmdEnv = append(cmdEnv, "GIT_PAGER=cat")
+	}
+	if !has("PAGER") {
+		cmdEnv = append(cmdEnv, "PAGER=cat")
+	}
+	if !has("SYSTEMD_PAGER") {
+		cmdEnv = append(cmdEnv, "SYSTEMD_PAGER=cat")
+	}
+	if !has("DEBIAN_FRONTEND") {
+		cmdEnv = append(cmdEnv, "DEBIAN_FRONTEND=noninteractive")
+	}
+	return cmdEnv
+}
+
+// attachNonInteractiveStdin 关闭交互式 stdin，使部分命令快速失败而非等待输入。
+func attachNonInteractiveStdin(cmd *exec.Cmd) {
+	if cmd == nil || cmd.Stdin != nil {
+		return
+	}
+	f, err := os.Open(os.DevNull)
+	if err != nil {
+		return
+	}
+	cmd.Stdin = f
+}

internal/security/shell_noninteractive_test.go

@@ -0,0 +1,128 @@
+package security
+
+import (
+	"context"
+	"os"
+	"os/exec"
+	"strings"
+	"testing"
+	"time"
+)
+
+func TestPrependNonInteractiveShellExports(t *testing.T) {
+	out := PrependNonInteractiveShellExports("echo hi")
+	if !strings.Contains(out, "GIT_PAGER=cat") || !strings.Contains(out, "PAGER=cat") {
+		t.Fatalf("missing pager exports: %q", out)
+	}
+	if !strings.HasSuffix(strings.TrimSpace(out), "echo hi") {
+		t.Fatalf("command suffix lost: %q", out)
+	}
+	skip := PrependNonInteractiveShellExports("GIT_PAGER=less echo hi")
+	if strings.Contains(skip, "export GIT_PAGER=cat") {
+		t.Fatalf("should not override existing GIT_PAGER: %q", skip)
+	}
+}
+
+func TestPrependNonInteractiveStdinRedirect(t *testing.T) {
+	out := PrependNonInteractiveStdinRedirect("echo hi")
+	if !strings.HasPrefix(out, "exec </dev/null") {
+		t.Fatalf("missing stdin redirect: %q", out)
+	}
+	if !strings.HasSuffix(strings.TrimSpace(out), "echo hi") {
+		t.Fatalf("command suffix lost: %q", out)
+	}
+	skip := PrependNonInteractiveStdinRedirect("cmd </dev/null")
+	if strings.HasPrefix(skip, "exec </dev/null") {
+		t.Fatalf("should not double redirect: %q", skip)
+	}
+}
+
+func TestPrepareNonInteractiveShellCommand(t *testing.T) {
+	out := PrepareNonInteractiveShellCommand("echo hi")
+	if !strings.Contains(out, "exec </dev/null") {
+		t.Fatalf("missing stdin redirect: %q", out)
+	}
+	if !strings.Contains(out, "GIT_PAGER=cat") {
+		t.Fatalf("missing pager export: %q", out)
+	}
+}
+
+func TestNewShellInactivityWatch(t *testing.T) {
+	w := NewShellInactivityWatch(1)
+	if w == nil {
+		t.Fatal("expected watch")
+	}
+	w.Bump()
+	select {
+	case <-w.Expired:
+	case <-time.After(3 * time.Second):
+		t.Fatal("expected inactivity fire within 3s")
+	}
+}
+
+func TestResolveShellNoOutputTimeoutSeconds(t *testing.T) {
+	if ResolveShellNoOutputTimeoutSeconds(0) != 300 {
+		t.Fatal("zero should default to 300")
+	}
+	if ResolveShellNoOutputTimeoutSeconds(-1) != 0 {
+		t.Fatal("-1 should disable")
+	}
+	if ResolveShellNoOutputTimeoutSeconds(30) != 30 {
+		t.Fatal("explicit value")
+	}
+}
+
+// TestNonInteractiveStdinReadExitsQuickly 验证 exec </dev/null + attachNonInteractiveStdin 时 read 立即 EOF，不挂起。
+func TestNonInteractiveStdinReadExitsQuickly(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping shell integration in -short")
+	}
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	cmd := exec.CommandContext(ctx, "sh", "-c", PrepareNonInteractiveShellCommand(`read x; echo "x=<$x>"`))
+	attachNonInteractiveStdin(cmd)
+
+	start := time.Now()
+	out, err := cmd.CombinedOutput()
+	elapsed := time.Since(start)
+	if elapsed > 2*time.Second {
+		t.Fatalf("read with closed stdin took %v, want <2s", elapsed)
+	}
+	if err != nil {
+		t.Fatalf("unexpected error: %v output=%q", err, out)
+	}
+	if !strings.Contains(string(out), "x=<>") {
+		t.Fatalf("unexpected output: %q", out)
+	}
+}
+
+// TestNonInteractiveStdinReadBlocksWithoutRedirect 对照：stdin 为永不写入的管道时 read 会挂起。
+func TestNonInteractiveStdinReadBlocksWithoutRedirect(t *testing.T) {
+	if testing.Short() {
+		t.Skip("skipping shell integration in -short")
+	}
+	r, w, err := os.Pipe()
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer r.Close()
+	// 保持 w 打开且不写数据，模拟「等待用户输入」
+
+	cmd := exec.Command("sh", "-c", `read x; echo done`)
+	cmd.Stdin = r
+
+	done := make(chan error, 1)
+	go func() { done <- cmd.Run() }()
+
+	select {
+	case err := <-done:
+		t.Fatalf("expected hang, but command finished: %v", err)
+	case <-time.After(500 * time.Millisecond):
+		if cmd.Process != nil {
+			_ = cmd.Process.Kill()
+		}
+		_ = w.Close()
+		<-done // 等待 goroutine 退出
+	}
+}

web/static/css/style.css

@@ -2456,16 +2456,68 @@ header {
     flex-shrink: 0;
 }
 
-.mcp-call-buttons {
+.mcp-call-buttons,
+.mcp-call-toolbar {
     display: flex;
     flex-wrap: wrap;
     gap: 6px;
+    align-items: center;
+}
+
+.mcp-tool-list {
+    display: none;
+    flex-wrap: wrap;
+    gap: 6px;
+    margin-top: 8px;
+    width: 100%;
+}
+
+.mcp-tool-list.expanded {
+    display: flex;
+}
+
+.mcp-tools-toggle-btn {
+    background: rgba(25, 118, 210, 0.1) !important;
+    border-color: rgba(25, 118, 210, 0.35) !important;
+    color: #1976d2 !important;
+}
+
+.mcp-call-toolbar .process-detail-btn,
+.mcp-call-toolbar .mcp-tools-toggle-btn {
+    display: inline-flex;
+    align-items: center;
+    justify-content: center;
+    gap: 6px;
+    min-height: 32px;
+    padding: 6px 12px;
+    font-size: 0.8125rem;
+    font-weight: 500;
+    line-height: 1.25;
+    box-sizing: border-box;
+    white-space: nowrap;
+    vertical-align: middle;
+}
+
+.mcp-call-toolbar .process-detail-btn span,
+.mcp-call-toolbar .mcp-tools-toggle-btn span {
+    display: inline-flex;
+    align-items: center;
+    line-height: 1.25;
+}
+
+.mcp-tools-toggle-btn:hover {
+    background: rgba(25, 118, 210, 0.18) !important;
+    border-color: #1976d2 !important;
+    color: #1565c0 !important;
 }
 
 .process-detail-btn {
     background: rgba(156, 39, 176, 0.1) !important;
     border-color: rgba(156, 39, 176, 0.3) !important;
     color: #9c27b0 !important;
+}
+
+.mcp-call-toolbar .process-detail-btn {
     display: inline-flex;
     align-items: center;
     gap: 6px;
@@ -11797,34 +11849,44 @@ tr.mcp-stats-tool-row[data-tool-name]:focus-visible {
     background: transparent;
     color: var(--text-muted);
     cursor: pointer;
-    border-radius: 4px;
+    border-radius: 6px;
     display: flex;
     align-items: center;
     justify-content: center;
     transition: all 0.2s ease;
 }
 
+.batch-delete-btn svg {
+    width: 16px;
+    height: 16px;
+    transition: transform 0.2s ease;
+}
+
 .batch-delete-btn:hover {
     background: rgba(220, 53, 69, 0.1);
     color: var(--error-color);
 }
 
+.batch-delete-btn:hover svg {
+    transform: scale(1.08);
+}
+
+.batch-delete-btn:active {
+    background: rgba(220, 53, 69, 0.2);
+    transform: scale(0.95);
+}
+
 .batch-manage-footer {
     display: flex;
     align-items: center;
-    justify-content: space-between;
+    justify-content: flex-end;
     padding: 16px 24px;
     border-top: 1px solid var(--border-color);
     flex-shrink: 0;
 }
 
-.select-all-checkbox {
+.batch-table-col-checkbox input[type="checkbox"] {
-    display: flex;
-    align-items: center;
-    gap: 8px;
     cursor: pointer;
-    font-size: 0.875rem;
-    color: var(--text-primary);
 }
 
 .batch-footer-actions {

web/static/i18n/en-US.json

@@ -530,6 +530,8 @@
     "noMatchTools": "No matching tools",
     "penetrationTestDetail": "Penetration test details",
     "expandDetail": "Expand details",
+    "toolExecutionsCount": "{{n}} tool runs",
+    "collapseToolExecutions": "Collapse tool runs",
     "noProcessDetail": "No process details (execution may be too fast or no detailed events)",
     "copyMessageTitle": "Copy message",
     "deleteTurnTitle": "Delete this turn",
@@ -1656,6 +1658,7 @@
     "rateWarning": "Some failures detected",
     "rateCritical": "High failure rate",
     "statsSubtitle": "Refreshed {{time}} · {{count}} tools",
+    "retentionHint": "Execution records are kept for {{days}} days, then purged automatically.",
     "timelineTitle": "Call trend",
     "timelineHint": "All tools combined (not split by tool)",
     "timelineRange24h": "24h",
@@ -2577,6 +2580,8 @@
     "agentModeSingle": "Single-agent (Eino ADK)",
     "agentModeMulti": "Multi-agent (Eino)",
     "agentModeHint": "Same as chat: Eino single-agent (ADK), or Deep / Plan-Execute / Supervisor (last three require multi_agent.enabled).",
+    "concurrency": "Concurrency",
+    "concurrencyHint": "Number of subtasks to run in parallel (1-8). Default 1 is serial; use 1-2 for scan-heavy tasks.",
     "scheduleMode": "Schedule mode",
     "scheduleModeManual": "Manual",
     "scheduleModeCron": "Cron expression",
@@ -2591,8 +2596,8 @@
     "tasksList": "Task list (one task per line)",
     "tasksListPlaceholder": "Enter task list, one per line",
     "tasksListPlaceholderExample": "Enter task list, one per line, for example:\nScan open ports of 192.168.1.1\nCheck if https://example.com has SQL injection\nEnumerate subdomains of example.com",
-    "tasksListHint": "Enter one task command per line; the system will execute them in order. Empty lines are ignored.",
+    "tasksListHint": "Enter one task command per line; the system runs them via a concurrency pool. Empty lines are ignored.",
-    "tasksListHintFull": "Hint: Enter one task command per line; the system will execute these tasks in order. Empty lines are ignored.",
+    "tasksListHintFull": "Hint: Enter one task command per line; the system runs them via a concurrency pool. Empty lines are ignored.",
     "createQueue": "Create queue"
   },
   "batchQueueDetailModal": {
@@ -2626,6 +2631,8 @@
     "scheduleToggleFailed": "Failed to update schedule toggle",
     "completedAt": "Completed at",
     "taskTotal": "Total tasks",
+    "concurrency": "Concurrency",
+    "concurrencyEditHint": "Click to edit. Cannot change while the queue is running.",
     "taskList": "Task list",
     "startLabel": "Start",
     "completeLabel": "Complete",

web/static/i18n/zh-CN.json

@@ -518,6 +518,8 @@
     "noMatchTools": "没有匹配的工具",
     "penetrationTestDetail": "渗透测试详情",
     "expandDetail": "展开详情",
+    "toolExecutionsCount": "{{n}}次工具执行",
+    "collapseToolExecutions": "收起工具执行",
     "noProcessDetail": "暂无过程详情（可能执行过快或未触发详细事件）",
     "copyMessageTitle": "复制消息内容",
     "deleteTurnTitle": "删除本轮对话",
@@ -1644,6 +1646,7 @@
     "rateWarning": "存在失败调用",
     "rateCritical": "失败率偏高",
     "statsSubtitle": "最后刷新 {{time}} · 共 {{count}} 个工具",
+    "retentionHint": "执行记录保留 {{days}} 天，超期自动清理",
     "timelineTitle": "调用趋势",
     "timelineHint": "全部工具合计，不按工具拆分",
     "timelineRange24h": "24 小时",
@@ -2565,6 +2568,8 @@
     "agentModeSingle": "单代理（Eino ADK）",
     "agentModeMulti": "多代理（Eino）",
     "agentModeHint": "与对话页一致：Eino 单代理（ADK），或 Deep / Plan-Execute / Supervisor（后三种需已启用多代理）。",
+    "concurrency": "并发数",
+    "concurrencyHint": "同时执行的子任务数量（1-8）。默认 1 为串行；含扫描类工具时建议 1-2。",
     "scheduleMode": "调度方式",
     "scheduleModeManual": "手工执行",
     "scheduleModeCron": "调度表达式（Cron）",
@@ -2579,8 +2584,8 @@
     "tasksList": "任务列表（每行一个任务）",
     "tasksListPlaceholder": "请输入任务列表，每行一个任务",
     "tasksListPlaceholderExample": "请输入任务列表，每行一个任务，例如：\n扫描 192.168.1.1 的开放端口\n检查 https://example.com 是否存在SQL注入\n枚举 example.com 的子域名",
-    "tasksListHint": "每行输入一个任务指令，系统将依次执行这些任务。空行会被自动忽略。",
+    "tasksListHint": "每行输入一个任务指令，系统将按并发池执行这些任务。空行会被自动忽略。",
-    "tasksListHintFull": "提示：每行输入一个任务指令，系统将依次执行这些任务。空行会被自动忽略。",
+    "tasksListHintFull": "提示：每行输入一个任务指令，系统将按并发池执行这些任务。空行会被自动忽略。",
     "createQueue": "创建队列"
   },
   "batchQueueDetailModal": {
@@ -2614,6 +2619,8 @@
     "scheduleToggleFailed": "更新调度开关失败",
     "completedAt": "完成时间",
     "taskTotal": "任务总数",
+    "concurrency": "并发数",
+    "concurrencyEditHint": "点击可修改；队列运行中不可改。",
     "taskList": "任务列表",
     "startLabel": "开始",
     "completeLabel": "完成",

web/static/js/chat.js

@@ -1,6 +1,39 @@
 let currentConversationId = null;
 let loadConversationRequestSeq = 0;
 
+/** 轻量会话 LRU 缓存：来回切换已加载会话时避免重复网络 + 全量 DOM 重建 */
+const CONVERSATION_LITE_CACHE_MAX = 12;
+const conversationLiteCache = new Map();
+
+function getConversationLiteFromCache(conversationId) {
+    if (!conversationId) return null;
+    const hit = conversationLiteCache.get(conversationId);
+    if (!hit) return null;
+    conversationLiteCache.delete(conversationId);
+    conversationLiteCache.set(conversationId, hit);
+    return hit;
+}
+
+function putConversationLiteCache(conversationId, data) {
+    if (!conversationId || !data) return;
+    conversationLiteCache.delete(conversationId);
+    conversationLiteCache.set(conversationId, data);
+    while (conversationLiteCache.size > CONVERSATION_LITE_CACHE_MAX) {
+        const oldest = conversationLiteCache.keys().next().value;
+        conversationLiteCache.delete(oldest);
+    }
+}
+
+function invalidateConversationLiteCache(conversationId) {
+    if (conversationId) {
+        conversationLiteCache.delete(conversationId);
+    } else {
+        conversationLiteCache.clear();
+    }
+}
+
+window.invalidateConversationLiteCache = invalidateConversationLiteCache;
+
 // @ 提及相关状态
 let mentionTools = [];
 let mentionToolsLoaded = false;
@@ -886,6 +919,9 @@ async function sendMessage() {
         window.CyberStrikeChatScroll.onUserSendMessage();
     }
     addMessage('user', displayMessage, null, null, null, { scroll: 'none' });
+    if (currentConversationId) {
+        invalidateConversationLiteCache(currentConversationId);
+    }
     
     // 清除防抖定时器，防止在清空输入框后重新保存草稿
     if (draftSaveTimer) {
@@ -2027,31 +2063,13 @@ function addMessage(role, content, mcpExecutionIds = null, progressId = null, cr
     
     // 有 MCP 执行记录且非流式占位消息时展示调用按钮；带 progressId 的流式占位不挂此条（与进度卡片一致，结束时 integrate 再创建）
     if (role === 'assistant' && (mcpExecutionIds && Array.isArray(mcpExecutionIds) && mcpExecutionIds.length > 0) && !progressId) {
-        const mcpSection = document.createElement('div');
+        if (options && options.deferMcpButtons) {
-        mcpSection.className = 'mcp-call-section';
+            try {
-        
+                messageDiv.dataset.pendingMcpExecutionIds = JSON.stringify(mcpExecutionIds);
-        const mcpLabel = document.createElement('div');
+            } catch (e) { /* ignore */ }
-        mcpLabel.className = 'mcp-call-label';
+        } else {
-        mcpLabel.textContent = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
+            appendMcpCallButtons(messageDiv, mcpExecutionIds);
-        mcpSection.appendChild(mcpLabel);
+        }
-        
-        const buttonsContainer = document.createElement('div');
-        buttonsContainer.className = 'mcp-call-buttons';
-        
-        mcpExecutionIds.forEach((execId, index) => {
-            const detailBtn = document.createElement('button');
-            detailBtn.className = 'mcp-detail-btn';
-            detailBtn.dataset.execId = execId;
-            detailBtn.dataset.execIndex = String(index + 1);
-            detailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.callNumber', { n: index + 1 }) : '调用 #' + (index + 1)) + '</span>';
-            detailBtn.onclick = () => showMCPDetail(execId);
-            buttonsContainer.appendChild(detailBtn);
-        });
-        // 使用批量 API 一次性获取所有工具名称（消除 N 次单独请求）
-        batchUpdateButtonToolNames(buttonsContainer, mcpExecutionIds);
-        
-        mcpSection.appendChild(buttonsContainer);
-        contentWrapper.appendChild(mcpSection);
     }
     
     messageDiv.appendChild(contentWrapper);
@@ -2151,11 +2169,13 @@ function copyMessageToClipboard(messageDiv, button) {
 function showCopySuccess(button) {
     if (button) {
         const originalText = button.innerHTML;
+        button.dataset.copySuccessActive = '1';
         button.innerHTML = '<svg width="16" height="16" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M20 6L9 17l-5-5" stroke="currentColor" stroke-width="2.5" stroke-linecap="round" stroke-linejoin="round" fill="none"/></svg><span>' + (typeof window.t === 'function' ? window.t('common.copied') : '已复制') + '</span>';
         button.style.color = '#10b981';
         button.style.background = 'rgba(16, 185, 129, 0.1)';
         button.style.borderColor = 'rgba(16, 185, 129, 0.3)';
         setTimeout(() => {
+            delete button.dataset.copySuccessActive;
             button.innerHTML = originalText;
             button.style.color = '';
             button.style.background = '';
@@ -2301,47 +2321,20 @@ function processDetailRowFingerprint(d) {
 }
 
 // 渲染过程详情
-function renderProcessDetails(messageId, processDetails) {
+// options.append=true 时分页追加；options.markLoaded=false 时保留 lazy 标记（分页加载中）
+function renderProcessDetails(messageId, processDetails, options) {
+    const renderOpts = options || {};
+    const appendMode = !!renderOpts.append;
+    const markLoaded = renderOpts.markLoaded !== false;
     const messageElement = document.getElementById(messageId);
     if (!messageElement) {
         return;
     }
     
-    // 查找或创建MCP调用区域
+    // 查找或创建 MCP 区域（工具栏 + 工具列表 + 迭代时间线 分区）
-    let mcpSection = messageElement.querySelector('.mcp-call-section');
+    const chrome = ensureMcpCallSectionChrome(messageElement, messageId);
-    if (!mcpSection) {
+    if (!chrome) return;
-        mcpSection = document.createElement('div');
+    const { mcpSection, toolbar: buttonsContainer } = chrome;
-        mcpSection.className = 'mcp-call-section';
-        
-        const contentWrapper = messageElement.querySelector('.message-content');
-        if (contentWrapper) {
-            contentWrapper.appendChild(mcpSection);
-        } else {
-            return;
-        }
-    }
-    
-    // 确保有标签和按钮容器（统一结构）
-    let mcpLabel = mcpSection.querySelector('.mcp-call-label');
-    let buttonsContainer = mcpSection.querySelector('.mcp-call-buttons');
-    
-    // 如果没有标签，创建一个（当没有工具调用时）
-    if (!mcpLabel && !buttonsContainer) {
-        mcpLabel = document.createElement('div');
-        mcpLabel.className = 'mcp-call-label';
-        mcpLabel.textContent = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
-        mcpSection.appendChild(mcpLabel);
-    } else if (mcpLabel && mcpLabel.textContent !== ('📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情'))) {
-        // 如果标签存在但不是统一格式，更新它
-        mcpLabel.textContent = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
-    }
-    
-    // 如果没有按钮容器，创建一个
-    if (!buttonsContainer) {
-        buttonsContainer = document.createElement('div');
-        buttonsContainer.className = 'mcp-call-buttons';
-        mcpSection.appendChild(buttonsContainer);
-    }
     
     // 添加过程详情按钮（如果还没有）
     let processDetailBtn = buttonsContainer.querySelector('.process-detail-btn');
@@ -2352,17 +2345,20 @@ function renderProcessDetails(messageId, processDetails) {
         processDetailBtn.onclick = () => toggleProcessDetails(null, messageId);
         buttonsContainer.appendChild(processDetailBtn);
     }
+    syncMcpToolsToggleButton(messageElement);
     
-    // 创建过程详情容器（放在按钮容器之后）
+    // 创建过程详情容器（放在工具列表之后）
     const detailsId = 'process-details-' + messageId;
     let detailsContainer = document.getElementById(detailsId);
+    const toolListEl = chrome.toolList;
     
     if (!detailsContainer) {
         detailsContainer = document.createElement('div');
         detailsContainer.id = detailsId;
         detailsContainer.className = 'process-details-container';
-        // 确保容器在按钮容器之后
+        if (toolListEl) {
-        if (buttonsContainer.nextSibling) {
+            toolListEl.after(detailsContainer);
+        } else if (buttonsContainer.nextSibling) {
             mcpSection.insertBefore(detailsContainer, buttonsContainer.nextSibling);
         } else {
             mcpSection.appendChild(detailsContainer);
@@ -2391,17 +2387,21 @@ function renderProcessDetails(messageId, processDetails) {
     if (isLazyNotLoaded && !reasoningFromMessage) {
         detailsContainer.dataset.lazyNotLoaded = '1';
         detailsContainer.dataset.loaded = '0';
-        timeline.innerHTML = '<div class="progress-timeline-empty">' +
+        const expandLabel = typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情';
-            (typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情') +
+        let lazyHint = expandLabel + '（点击后加载迭代详情）';
-            '（点击后加载）</div>';
+        timeline.innerHTML = '<div class="progress-timeline-empty">' + lazyHint + '</div>';
         timeline.classList.remove('expanded');
+        prefetchProcessDetailsSummaryHint(messageId, messageElement);
         return;
     }
     if (isLazyNotLoaded) {
         detailsContainer.dataset.lazyNotLoaded = '1';
         detailsContainer.dataset.loaded = '0';
         processDetails = [];
-    } else {
+        if (!appendMode) {
+            prefetchProcessDetailsSummaryHint(messageId, messageElement);
+        }
+    } else if (markLoaded) {
         detailsContainer.dataset.lazyNotLoaded = '0';
         detailsContainer.dataset.loaded = '1';
     }
@@ -2413,15 +2413,16 @@ function renderProcessDetails(messageId, processDetails) {
     }
     // 如果没有processDetails或为空，显示空状态
     if (!processDetails || processDetails.length === 0) {
-        // 显示空状态提示
+        if (!appendMode) {
             timeline.innerHTML = '<div class="progress-timeline-empty">' + (typeof window.t === 'function' ? window.t('chat.noProcessDetail') : '暂无过程详情（可能执行过快或未触发详细事件）') + '</div>';
-        // 默认折叠
             timeline.classList.remove('expanded');
+        }
         return;
     }
     
-    // 清空时间线并重新渲染
+    if (!appendMode) {
         timeline.innerHTML = '';
+    }
     
     
     function processDetailAgentPrefix(d) {
@@ -2430,14 +2431,12 @@ function renderProcessDetails(messageId, processDetails) {
         return s ? ('[' + s + '] ') : '';
     }
 
-    // 渲染每个过程详情事件
+    function renderOneProcessDetail(detail) {
-    processDetails.forEach(detail => {
         const eventType = detail.eventType || '';
         const title = detail.message || '';
         const data = detail.data || {};
         const agPx = processDetailAgentPrefix(data);
         
-        // 根据事件类型渲染不同的内容
         let itemTitle = title;
         if (eventType === 'iteration') {
             const n = data.iteration || 1;
@@ -2530,15 +2529,38 @@ function renderProcessDetails(messageId, processDetails) {
             title: itemTitle,
             message: detail.message || '',
             data: data,
-            createdAt: detail.createdAt // 传递实际的事件创建时间
+            createdAt: detail.createdAt
         };
         if (eventType === 'tool_call' && data._mergedResult) {
             timelineOpts.mergedResult = data._mergedResult;
         }
         addTimelineItem(timeline, eventType, timelineOpts);
-    });
+    }
 
-    if (isLazyNotLoaded && reasoningFromMessage) {
+    const TIMELINE_RENDER_BATCH = 40;
+    const renderTimelineBatch = (startIdx) => {
+        const endIdx = Math.min(startIdx + TIMELINE_RENDER_BATCH, processDetails.length);
+        for (let i = startIdx; i < endIdx; i++) {
+            renderOneProcessDetail(processDetails[i]);
+        }
+        if (endIdx < processDetails.length) {
+            requestAnimationFrame(() => renderTimelineBatch(endIdx));
+        } else if (markLoaded) {
+            finishProcessDetailsRender(messageElement, processDetails, isLazyNotLoaded, timeline);
+        }
+    };
+    if (processDetails.length > TIMELINE_RENDER_BATCH) {
+        renderTimelineBatch(0);
+    } else {
+        processDetails.forEach(renderOneProcessDetail);
+        if (markLoaded) {
+            finishProcessDetailsRender(messageElement, processDetails, isLazyNotLoaded, timeline);
+        }
+    }
+}
+
+function finishProcessDetailsRender(messageElement, processDetails, isLazyNotLoaded, timeline) {
+    if (isLazyNotLoaded && getMessageReasoningContent(messageElement)) {
         const lazyHint = document.createElement('div');
         lazyHint.className = 'progress-timeline-empty progress-timeline-lazy-hint';
         lazyHint.textContent = (typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情') +
@@ -2546,15 +2568,12 @@ function renderProcessDetails(messageId, processDetails) {
         timeline.appendChild(lazyHint);
     }
     
-    // 检查是否有错误或取消事件，如果有，确保详情默认折叠（但仍有待审批 HITL 时保持展开，由 restoreHitlInlineForConversation 处理）
     const hasPendingHitlInDetails = processDetails.some(d => d && d.eventType === 'hitl_interrupt');
     const hasErrorOrCancelled = processDetails.some(d => 
         d.eventType === 'error' || d.eventType === 'cancelled'
     );
     if (hasErrorOrCancelled && !hasPendingHitlInDetails) {
-        // 确保时间线是折叠的
         timeline.classList.remove('expanded');
-        // 更新按钮文本为"展开详情"
         const processDetailBtn = messageElement.querySelector('.process-detail-btn');
         if (processDetailBtn) {
             processDetailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情') + '</span>';
@@ -2562,6 +2581,36 @@ function renderProcessDetails(messageId, processDetails) {
     }
 }
 
+/** 懒加载折叠态：后台拉摘要，提示迭代规模而不加载全量详情 */
+function prefetchProcessDetailsSummaryHint(messageId, messageElement) {
+    if (!messageElement || !messageElement.dataset || !messageElement.dataset.backendMessageId) return;
+    const backendId = String(messageElement.dataset.backendMessageId).trim();
+    if (!backendId || typeof apiFetch !== 'function') return;
+    const detailsContainer = document.getElementById('process-details-' + messageId);
+    if (!detailsContainer || detailsContainer.dataset.summaryFetched === '1') return;
+    detailsContainer.dataset.summaryFetched = '1';
+    apiFetch('/api/messages/' + encodeURIComponent(backendId) + '/process-details?summary=1')
+        .then(async (res) => {
+            const j = await res.json().catch(() => ({}));
+            if (!res.ok || !j.summary) return;
+            const s = j.summary;
+            const timeline = detailsContainer.querySelector('.progress-timeline');
+            if (!timeline || detailsContainer.dataset.loaded === '1') return;
+            const expandLabel = typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情';
+            let hint = expandLabel + '（点击后加载迭代详情）';
+            if (s.maxIteration > 0) {
+                hint = expandLabel + '（共 ' + s.maxIteration + ' 轮迭代，' + (s.total || 0) + ' 条详情）';
+            } else if (s.total > 0) {
+                hint = expandLabel + '（共 ' + (s.total || 0) + ' 条详情）';
+            }
+            const empty = timeline.querySelector('.progress-timeline-empty');
+            if (empty) {
+                empty.textContent = hint;
+            }
+        })
+        .catch(() => {});
+}
+
 // 移除消息
 function removeMessage(id) {
     const messageDiv = document.getElementById(id);
@@ -2623,6 +2672,201 @@ async function updateButtonWithToolName(button, executionId, index) {
     }
 }
 
+function getPendingMcpExecutionCount(messageElement) {
+    if (!messageElement || !messageElement.dataset || !messageElement.dataset.pendingMcpExecutionIds) {
+        return 0;
+    }
+    try {
+        const ids = JSON.parse(messageElement.dataset.pendingMcpExecutionIds);
+        return Array.isArray(ids) ? ids.length : 0;
+    } catch (e) {
+        return 0;
+    }
+}
+
+function getMcpExecutionCount(messageElement) {
+    const pending = getPendingMcpExecutionCount(messageElement);
+    if (pending > 0) return pending;
+    const toolList = messageElement && messageElement.querySelector('.mcp-tool-list');
+    if (toolList) {
+        return toolList.querySelectorAll('.mcp-detail-btn[data-exec-id]').length;
+    }
+    return 0;
+}
+
+function formatMcpToolsToggleLabel(count, expanded) {
+    if (expanded) {
+        if (typeof window.t === 'function') {
+            const s = window.t('chat.collapseToolExecutions');
+            if (s && s !== 'chat.collapseToolExecutions') return s;
+        }
+        return '收起工具执行';
+    }
+    if (typeof window.t === 'function') {
+        const s = window.t('chat.toolExecutionsCount', { n: count });
+        if (s && s !== 'chat.toolExecutionsCount') return s;
+    }
+    return count + '次工具执行';
+}
+
+/** 渗透测试区：工具栏（展开详情 | N次工具执行）+ 独立工具列表 + 迭代时间线 */
+function ensureMcpCallSectionChrome(messageElement, messageId) {
+    const contentWrapper = messageElement && messageElement.querySelector('.message-content');
+    if (!contentWrapper) return null;
+
+    let mcpSection = messageElement.querySelector('.mcp-call-section');
+    if (!mcpSection) {
+        mcpSection = document.createElement('div');
+        mcpSection.className = 'mcp-call-section';
+        const mcpLabel = document.createElement('div');
+        mcpLabel.className = 'mcp-call-label';
+        mcpLabel.textContent = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
+        mcpSection.appendChild(mcpLabel);
+        contentWrapper.appendChild(mcpSection);
+    } else {
+        const mcpLabel = mcpSection.querySelector('.mcp-call-label');
+        const labelText = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
+        if (mcpLabel && mcpLabel.textContent !== labelText) {
+            mcpLabel.textContent = labelText;
+        }
+    }
+
+    let toolbar = mcpSection.querySelector('.mcp-call-toolbar');
+    const legacyButtons = mcpSection.querySelector('.mcp-call-buttons');
+    if (!toolbar) {
+        toolbar = document.createElement('div');
+        toolbar.className = 'mcp-call-toolbar';
+        if (legacyButtons) {
+            const processBtn = legacyButtons.querySelector('.process-detail-btn');
+            if (processBtn) toolbar.appendChild(processBtn);
+            mcpSection.replaceChild(toolbar, legacyButtons);
+        } else {
+            mcpSection.appendChild(toolbar);
+        }
+    }
+
+    let toolList = mcpSection.querySelector('.mcp-tool-list');
+    if (!toolList) {
+        toolList = document.createElement('div');
+        toolList.className = 'mcp-tool-list';
+        const detailsContainer = mcpSection.querySelector('.process-details-container');
+        if (detailsContainer) {
+            mcpSection.insertBefore(toolList, detailsContainer);
+        } else {
+            toolbar.after(toolList);
+        }
+    }
+
+    if (legacyButtons && legacyButtons.parentNode === mcpSection) {
+        legacyButtons.querySelectorAll('.mcp-detail-btn[data-exec-id]').forEach((btn) => toolList.appendChild(btn));
+        legacyButtons.remove();
+    }
+
+    const clientId = messageId || messageElement.id;
+    if (clientId && !toolbar.querySelector('.process-detail-btn')) {
+        const processDetailBtn = document.createElement('button');
+        processDetailBtn.className = 'mcp-detail-btn process-detail-btn';
+        processDetailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情') + '</span>';
+        processDetailBtn.onclick = () => toggleProcessDetails(null, clientId);
+        toolbar.appendChild(processDetailBtn);
+    }
+
+    return { mcpSection, toolbar, toolList };
+}
+
+function syncMcpToolsToggleButton(messageElement) {
+    if (!messageElement) return;
+    const chrome = ensureMcpCallSectionChrome(messageElement, messageElement.id);
+    if (!chrome) return;
+    const { toolbar, toolList } = chrome;
+    const count = getMcpExecutionCount(messageElement);
+    let toolsToggle = toolbar.querySelector('.mcp-tools-toggle-btn');
+    if (count <= 0) {
+        if (toolsToggle) toolsToggle.remove();
+        return;
+    }
+    if (!toolsToggle) {
+        toolsToggle = document.createElement('button');
+        toolsToggle.type = 'button';
+        toolsToggle.className = 'mcp-detail-btn mcp-tools-toggle-btn';
+        toolsToggle.onclick = function (e) {
+            e.stopPropagation();
+            toggleMcpToolList(messageElement.id);
+        };
+        toolbar.appendChild(toolsToggle);
+    }
+    const expanded = toolList.classList.contains('expanded');
+    toolsToggle.innerHTML = '<span>' + formatMcpToolsToggleLabel(count, expanded) + '</span>';
+}
+
+function toggleMcpToolList(assistantMessageId) {
+    const messageEl = document.getElementById(assistantMessageId);
+    if (!messageEl) return;
+    const chrome = ensureMcpCallSectionChrome(messageEl, assistantMessageId);
+    if (!chrome) return;
+    const { toolList } = chrome;
+    const willExpand = !toolList.classList.contains('expanded');
+    if (willExpand) {
+        ensureMcpCallButtons(messageEl);
+        toolList.classList.add('expanded');
+    } else {
+        toolList.classList.remove('expanded');
+    }
+    syncMcpToolsToggleButton(messageEl);
+}
+
+window.toggleMcpToolList = toggleMcpToolList;
+window.syncMcpToolsToggleButton = syncMcpToolsToggleButton;
+window.ensureMcpCallSectionChrome = ensureMcpCallSectionChrome;
+
+/** 将 MCP 工具按钮挂到独立工具列表，并批量解析工具名 */
+function appendMcpCallButtons(messageElement, executionIds) {
+    if (!messageElement || !Array.isArray(executionIds) || executionIds.length === 0) {
+        return;
+    }
+    const chrome = ensureMcpCallSectionChrome(messageElement, messageElement.id);
+    if (!chrome) return;
+    const toolList = chrome.toolList;
+
+    executionIds.forEach((execId, index) => {
+        if (toolList.querySelector('.mcp-detail-btn[data-exec-id="' + CSS.escape(String(execId)) + '"]')) {
+            return;
+        }
+        const detailBtn = document.createElement('button');
+        detailBtn.className = 'mcp-detail-btn';
+        detailBtn.dataset.execId = execId;
+        detailBtn.dataset.execIndex = String(index + 1);
+        detailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.callNumber', { n: index + 1 }) : '调用 #' + (index + 1)) + '</span>';
+        detailBtn.onclick = () => showMCPDetail(execId);
+        toolList.appendChild(detailBtn);
+    });
+    batchUpdateButtonToolNames(toolList, executionIds);
+    syncMcpToolsToggleButton(messageElement);
+}
+
+/** 历史会话懒加载：用户展开工具列表时再渲染工具按钮 */
+function ensureMcpCallButtons(messageElement) {
+    if (!messageElement || !messageElement.dataset || !messageElement.dataset.pendingMcpExecutionIds) {
+        return;
+    }
+    let executionIds;
+    try {
+        executionIds = JSON.parse(messageElement.dataset.pendingMcpExecutionIds);
+    } catch (e) {
+        delete messageElement.dataset.pendingMcpExecutionIds;
+        return;
+    }
+    if (!Array.isArray(executionIds) || executionIds.length === 0) {
+        delete messageElement.dataset.pendingMcpExecutionIds;
+        return;
+    }
+    appendMcpCallButtons(messageElement, executionIds);
+    delete messageElement.dataset.pendingMcpExecutionIds;
+}
+
+window.ensureMcpCallButtons = ensureMcpCallButtons;
+window.appendMcpCallButtons = appendMcpCallButtons;
+
 // 批量获取工具名称并更新按钮（消除 N 次单独 API 请求，合并为 1 次）
 async function batchUpdateButtonToolNames(buttonsContainer, executionIds) {
     if (!executionIds || executionIds.length === 0) return;
@@ -2866,7 +3110,17 @@ async function cancelMCPToolExecutionSubmit(executionId, userNote, options = {})
     if (!executionId) {
         return;
     }
+    let conversationId = '';
+    if (typeof monitorState !== 'undefined' && Array.isArray(monitorState.executions)) {
+        const exec = monitorState.executions.find(e => e && e.id === executionId);
+        if (exec) {
+            conversationId = (exec.conversationId || '').trim();
+        }
+    }
     try {
+        if (conversationId && typeof requestCancelWithContinue === 'function') {
+            await requestCancelWithContinue(conversationId, userNote || '');
+        } else {
             const res = await apiFetch(`/api/monitor/execution/${encodeURIComponent(executionId)}/cancel`, {
                 method: 'POST',
                 headers: { 'Content-Type': 'application/json' },
@@ -2876,6 +3130,7 @@ async function cancelMCPToolExecutionSubmit(executionId, userNote, options = {})
             if (!res.ok) {
                 throw new Error(body.error || body.message || res.statusText);
             }
+        }
         const okMsg = typeof window.t === 'function' ? window.t('mcpDetailModal.abortSuccess') : '已发送终止请求';
         alert(okMsg);
         if (options.refreshDetail && typeof showMCPDetail === 'function') {
@@ -2892,7 +3147,7 @@ async function cancelMCPToolExecutionSubmit(executionId, userNote, options = {})
 }
 
 /**
- * 取消单次 MCP 工具执行（监控页「终止」）。弹出说明框后提交；仅取消该次 tools/call，不停止整条对话/迭代任务。
+ * 取消单次 MCP 工具执行（监控页「终止」）。有 conversationId 时复用对话页「中断并继续」弹窗与 API。
  * @param {string} executionId
  * @param {{ refreshDetail?: boolean }} [options]
  */
@@ -2900,6 +3155,18 @@ async function cancelMCPToolExecution(executionId, options = {}) {
     if (!executionId) {
         return;
     }
+    let conversationId = '';
+    if (typeof monitorState !== 'undefined' && Array.isArray(monitorState.executions)) {
+        const exec = monitorState.executions.find(e => e && e.id === executionId);
+        if (exec) {
+            conversationId = (exec.conversationId || '').trim();
+        }
+    }
+    if (conversationId && typeof openUserInterruptModal === 'function') {
+        openUserInterruptModal(null, conversationId);
+        window.__monitorInterruptContext = { executionId: executionId, options: options || {} };
+        return;
+    }
     openMcpToolAbortModal(executionId, options);
 }
 
@@ -3182,41 +3449,64 @@ function getConversationGroup(dateObj, todayStart, sevenDaysCutoff, yesterdaySta
 }
 
 // 加载对话
-/** 轻量加载会话后，拉取最后一条助手消息的 process_details（机器人等无 SSE 场景） */
+/** 轻量加载会话后，仅对「处理中…」占位回复拉取过程详情（机器人等非 SSE 场景）；已完成会话不预取全量 */
 async function prefetchLastAssistantProcessDetails() {
     const nodes = document.querySelectorAll('#chat-messages .message.assistant');
     if (!nodes.length) return;
     const last = nodes[nodes.length - 1];
     if (!last || !last.id) return;
+    const bubble = last.querySelector('.message-bubble');
+    const visibleText = bubble ? String(bubble.textContent || '').trim() : '';
+    const isPlaceholder = visibleText === '处理中...' || visibleText === 'Processing...';
+    if (!isPlaceholder) return;
     const container = document.getElementById('process-details-' + last.id);
     if (!container || container.dataset.lazyNotLoaded !== '1') return;
     const backendId = last.dataset && last.dataset.backendMessageId;
     if (!backendId || typeof apiFetch !== 'function') return;
+    if (typeof window.loadProcessDetailsPaginated === 'function') {
+        await window.loadProcessDetailsPaginated(last.id, backendId);
+        return;
+    }
     const res = await apiFetch('/api/messages/' + encodeURIComponent(String(backendId)) + '/process-details');
     const j = await res.json().catch(() => ({}));
     if (!res.ok || !Array.isArray(j.processDetails) || j.processDetails.length === 0) return;
     if (typeof renderProcessDetails === 'function') {
         renderProcessDetails(last.id, j.processDetails);
     }
-    if (typeof window.expandProcessDetailsTimeline === 'function') {
-        window.expandProcessDetailsTimeline(last.id);
-    }
 }
 
 async function loadConversation(conversationId) {
     const seq = ++loadConversationRequestSeq;
     try {
-        // 轻量加载：不带 processDetails，避免历史会话切换卡顿；展开详情时再按需拉取
+        const cachedConversation = getConversationLiteFromCache(conversationId);
-        const response = await apiFetch(`/api/conversations/${conversationId}?include_process_details=0`);
+        const fetchPromise = apiFetch(`/api/conversations/${conversationId}?include_process_details=0`)
+            .then(async (response) => {
+                const data = await response.json();
+                return { response, data };
+            });
+
+        let conversation;
+        let response;
+        if (cachedConversation) {
+            conversation = cachedConversation;
+            fetchPromise.then(({ response: freshResp, data }) => {
+                if (freshResp.ok && data && seq === loadConversationRequestSeq && currentConversationId === conversationId) {
+                    putConversationLiteCache(conversationId, data);
+                }
+            }).catch(() => {});
+        } else {
+            const fetched = await fetchPromise;
+            response = fetched.response;
+            conversation = fetched.data;
             if (seq !== loadConversationRequestSeq) {
                 return;
             }
-        const conversation = await response.json();
-        
             if (!response.ok) {
                 showChatToast('加载对话失败: ' + (conversation.error || '未知错误'), 'error');
                 return;
             }
+            putConversationLiteCache(conversationId, conversation);
+        }
         if (seq !== loadConversationRequestSeq) {
             return;
         }
@@ -3265,11 +3555,15 @@ async function loadConversation(conversationId) {
         if (typeof refreshChatProjectSelector === 'function') {
             refreshChatProjectSelector();
         }
-        if (typeof window.syncHitlConfigFromServer === 'function') {
+        refreshHitlConfigByCurrentConversation();
-            await window.syncHitlConfigFromServer(conversationId);
+        const hitlSyncPromise = (typeof window.syncHitlConfigFromServer === 'function')
-        } else {
+            ? window.syncHitlConfigFromServer(conversationId).then(() => {
+                if (seq === loadConversationRequestSeq && currentConversationId === conversationId) {
                     refreshHitlConfigByCurrentConversation();
                 }
+            }).catch(() => {})
+            : Promise.resolve();
+        void hitlSyncPromise;
         updateActiveConversation();
         
         // 如果攻击链模态框打开且显示的不是当前对话，关闭它
@@ -3336,7 +3630,9 @@ async function loadConversation(conversationId) {
                 // - user: createdAt 即可（发送后不会再更新）
                 // - assistant: 如果后端提供 updatedAt（任务完成时写回），优先用它，避免占位消息“任务开始时间”误导
                 const msgTime = (msg && msg.role === 'assistant' && msg.updatedAt) ? msg.updatedAt : (msg ? msg.createdAt : null);
-                const messageId = addMessage(msg.role, displayContent, msg.mcpExecutionIds || [], null, msgTime);
+                const mcpIds = (msg.mcpExecutionIds && Array.isArray(msg.mcpExecutionIds)) ? msg.mcpExecutionIds : [];
+                const addOpts = (msg.role === 'assistant' && mcpIds.length > 0) ? { deferMcpButtons: true } : null;
+                const messageId = addMessage(msg.role, displayContent, mcpIds, null, msgTime, addOpts);
                 const messageEl = document.getElementById(messageId);
                 if (messageEl && msg && msg.id) {
                     messageEl.dataset.backendMessageId = String(msg.id);
@@ -3504,6 +3800,7 @@ async function deleteConversationTurnFromUI(anchorBackendMessageId) {
         if (!response.ok) {
             throw new Error(data.error || data.message || 'delete failed');
         }
+        invalidateConversationLiteCache(currentConversationId);
         await loadConversation(currentConversationId);
         if (typeof loadConversationsWithGroups === 'function') {
             loadConversationsWithGroups();
@@ -3550,6 +3847,7 @@ async function deleteConversation(conversationId, skipConfirm = false) {
         
         // 更新缓存 - 立即删除，确保后续加载时能正确识别
         delete conversationGroupMappingCache[conversationId];
+        invalidateConversationLiteCache(conversationId);
         // 同时从待保留映射中移除
         delete pendingGroupMappings[conversationId];
         
@@ -7450,14 +7748,14 @@ async function showBatchManageModal() {
         updateBatchManageTitle(allConversationsForBatch.length);
 
         renderBatchConversations();
-        openAppModal('batch-manage-modal');
+        openAppModal('batch-manage-modal', { focus: false });
     } catch (error) {
         console.error('加载对话列表失败:', error);
         // 错误时使用空数组，不显示错误提示（更友好的用户体验）
         allConversationsForBatch = [];
         updateBatchManageTitle(0);
         renderBatchConversations();
-        openAppModal('batch-manage-modal');
+        openAppModal('batch-manage-modal', { focus: false });
     }
 }
 
@@ -7517,6 +7815,7 @@ function renderBatchConversations(filtered = null) {
         checkbox.type = 'checkbox';
         checkbox.className = 'batch-conversation-checkbox';
         checkbox.dataset.conversationId = conv.id;
+        checkbox.addEventListener('change', syncSelectAllBatchCheckbox);
 
         const name = document.createElement('div');
         name.className = 'batch-table-col-name';
@@ -7542,9 +7841,21 @@ function renderBatchConversations(filtered = null) {
         const action = document.createElement('div');
         action.className = 'batch-table-col-action';
         const deleteBtn = document.createElement('button');
+        deleteBtn.type = 'button';
         deleteBtn.className = 'batch-delete-btn';
-        deleteBtn.innerHTML = '🗑️';
+        deleteBtn.innerHTML = `
-        deleteBtn.onclick = () => deleteConversation(conv.id);
+            <svg width="16" height="16" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg" aria-hidden="true">
+                <path d="M3 6h18M8 6V4a2 2 0 0 1 2-2h4a2 2 0 0 1 2 2v2m3 0v14a2 2 0 0 1-2 2H7a2 2 0 0 1-2-2V6h14zM10 11v6M14 11v6"
+                      stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"/>
+            </svg>
+        `;
+        const deleteLabel = typeof window.t === 'function' ? window.t('contextMenu.deleteConversation') : '删除此对话';
+        deleteBtn.title = deleteLabel;
+        deleteBtn.setAttribute('aria-label', deleteLabel);
+        deleteBtn.onclick = (e) => {
+            e.stopPropagation();
+            deleteConversation(conv.id);
+        };
         action.appendChild(deleteBtn);
 
         row.appendChild(checkbox);
@@ -7554,6 +7865,8 @@ function renderBatchConversations(filtered = null) {
 
         list.appendChild(row);
     });
+
+    syncSelectAllBatchCheckbox();
 }
 
 // 筛选批量管理对话
@@ -7576,11 +7889,34 @@ function toggleSelectAllBatch() {
     const selectAll = document.getElementById('batch-select-all');
     const checkboxes = document.querySelectorAll('.batch-conversation-checkbox');
 
+    if (selectAll) {
+        selectAll.indeterminate = false;
+    }
     checkboxes.forEach(cb => {
         cb.checked = selectAll.checked;
     });
 }
 
+function syncSelectAllBatchCheckbox() {
+    const selectAll = document.getElementById('batch-select-all');
+    if (!selectAll) return;
+
+    const checkboxes = document.querySelectorAll('.batch-conversation-checkbox');
+    const total = checkboxes.length;
+    const checked = document.querySelectorAll('.batch-conversation-checkbox:checked').length;
+
+    if (total === 0 || checked === 0) {
+        selectAll.checked = false;
+        selectAll.indeterminate = false;
+    } else if (checked === total) {
+        selectAll.checked = true;
+        selectAll.indeterminate = false;
+    } else {
+        selectAll.checked = false;
+        selectAll.indeterminate = true;
+    }
+}
+
 // 删除选中的对话
 async function deleteSelectedConversations() {
     const checkboxes = document.querySelectorAll('.batch-conversation-checkbox:checked');
@@ -7604,6 +7940,7 @@ async function deleteSelectedConversations() {
         const selectAll = document.getElementById('batch-select-all');
         if (selectAll) {
             selectAll.checked = false;
+            selectAll.indeterminate = false;
         }
     } catch (error) {
         console.error('删除失败:', error);
@@ -7619,6 +7956,7 @@ function closeBatchManageModal() {
     const selectAll = document.getElementById('batch-select-all');
     if (selectAll) {
         selectAll.checked = false;
+        selectAll.indeterminate = false;
     }
     allConversationsForBatch = [];
 }
@@ -7653,6 +7991,20 @@ function refreshChatPanelI18n() {
             const expanded = timeline && timeline.classList.contains('expanded');
             span.textContent = expanded ? t('tasks.collapseDetail') : t('chat.expandDetail');
         });
+        const copyLabel = t('common.copy');
+        const copyTitle = t('chat.copyMessageTitle');
+        messagesEl.querySelectorAll('.message-copy-btn').forEach(function (btn) {
+            if (btn.dataset.copySuccessActive === '1') return;
+            const span = btn.querySelector('span');
+            if (span) span.textContent = copyLabel;
+            btn.title = copyTitle;
+            btn.setAttribute('aria-label', copyTitle);
+        });
+        messagesEl.querySelectorAll('.message.assistant').forEach(function (msgEl) {
+            if (typeof window.syncMcpToolsToggleButton === 'function') {
+                window.syncMcpToolsToggleButton(msgEl);
+            }
+        });
     }
 
     if (isAppModalOpen('mcp-detail-modal')) {

web/static/js/monitor.js

@@ -1003,6 +1003,7 @@ function openUserInterruptModal(progressId, conversationId) {
 
 function closeUserInterruptModal() {
     userInterruptModalPending = null;
+    window.__monitorInterruptContext = null;
     closeAppModal('user-interrupt-modal');
 }
 
@@ -1012,6 +1013,7 @@ async function submitUserInterruptContinue() {
     }
     const reason = (document.getElementById('user-interrupt-reason') && document.getElementById('user-interrupt-reason').value || '').trim();
     const { progressId, conversationId } = userInterruptModalPending;
+    const monitorCtx = window.__monitorInterruptContext;
     closeUserInterruptModal();
     const stopBtn = progressId ? document.getElementById(`${progressId}-stop-btn`) : null;
     try {
@@ -1020,6 +1022,13 @@ async function submitUserInterruptContinue() {
             stopBtn.textContent = typeof window.t === 'function' ? window.t('tasks.interruptSubmitting') : '提交中...';
         }
         await requestCancelWithContinue(conversationId, reason);
+        if (monitorCtx && monitorCtx.executionId && typeof refreshMonitorPanel === 'function') {
+            const page = (typeof monitorState !== 'undefined' && monitorState.pagination && monitorState.pagination.page)
+                ? monitorState.pagination.page
+                : 1;
+            await refreshMonitorPanel(page);
+            window.__monitorInterruptContext = null;
+        }
         loadActiveTasks();
     } catch (error) {
         console.error('中断并继续失败:', error);
@@ -1259,101 +1268,60 @@ function integrateProgressToMCPSection(progressId, assistantMessageId, mcpExecut
         return;
     }
     
-    // 查找或创建 MCP 区域
+    // 查找或创建 MCP 区域（工具栏 + 工具列表 + 迭代时间线）
-    let mcpSection = assistantElement.querySelector('.mcp-call-section');
+    if (typeof window.ensureMcpCallSectionChrome === 'function') {
+        window.ensureMcpCallSectionChrome(assistantElement, assistantMessageId);
+    }
+    const mcpSection = assistantElement.querySelector('.mcp-call-section');
     if (!mcpSection) {
-        mcpSection = document.createElement('div');
+        removeMessage(progressId);
-        mcpSection.className = 'mcp-call-section';
+        return;
-        const mcpLabel = document.createElement('div');
-        mcpLabel.className = 'mcp-call-label';
-        mcpLabel.textContent = '📋 ' + (typeof window.t === 'function' ? window.t('chat.penetrationTestDetail') : '渗透测试详情');
-        mcpSection.appendChild(mcpLabel);
-        const buttonsContainerInit = document.createElement('div');
-        buttonsContainerInit.className = 'mcp-call-buttons';
-        mcpSection.appendChild(buttonsContainerInit);
-        contentWrapper.appendChild(mcpSection);
     }
 
-    // 获取时间线内容
     const hasContent = timelineHTML.trim().length > 0;
 
-    // 检查时间线中是否有错误项
+    if (mcpIds.length > 0 && typeof window.appendMcpCallButtons === 'function') {
-    const hasError = timeline && timeline.querySelector('.timeline-item-error');
+        window.appendMcpCallButtons(assistantElement, mcpIds);
-    
+        const toolList = mcpSection.querySelector('.mcp-tool-list');
-    // 确保按钮容器存在
+        if (toolList) toolList.classList.remove('expanded');
-    let buttonsContainer = mcpSection.querySelector('.mcp-call-buttons');
+    }
-    if (!buttonsContainer) {
+    if (typeof window.syncMcpToolsToggleButton === 'function') {
-        buttonsContainer = document.createElement('div');
+        window.syncMcpToolsToggleButton(assistantElement);
-        buttonsContainer.className = 'mcp-call-buttons';
-        mcpSection.appendChild(buttonsContainer);
     }
 
-    let maxExecIndex = 0;
+    const toolbar = mcpSection.querySelector('.mcp-call-toolbar');
-    const existingExecBtns = buttonsContainer.querySelectorAll('.mcp-detail-btn:not(.process-detail-btn)');
+    if (toolbar && !toolbar.querySelector('.process-detail-btn')) {
-    existingExecBtns.forEach(function (btn) {
-        const n = parseInt(btn.dataset.execIndex, 10);
-        if (!isNaN(n) && n > maxExecIndex) maxExecIndex = n;
-    });
-    const seenExec = new Set();
-    existingExecBtns.forEach(function (btn) {
-        if (btn.dataset.execId) seenExec.add(String(btn.dataset.execId).trim());
-    });
-    let appendedAny = false;
-    if (mcpIds.length > 0) {
-        mcpIds.forEach(function (execId) {
-            const id = execId != null ? String(execId).trim() : '';
-            if (!id || seenExec.has(id)) return;
-            seenExec.add(id);
-            maxExecIndex += 1;
-            appendedAny = true;
-            const detailBtn = document.createElement('button');
-            detailBtn.className = 'mcp-detail-btn';
-            detailBtn.dataset.execId = id;
-            detailBtn.dataset.execIndex = String(maxExecIndex);
-            detailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.callNumber', { n: maxExecIndex }) : '调用 #' + maxExecIndex) + '</span>';
-            detailBtn.onclick = function () { showMCPDetail(id); };
-            buttonsContainer.appendChild(detailBtn);
-        });
-        if (appendedAny && typeof batchUpdateButtonToolNames === 'function') {
-            batchUpdateButtonToolNames(buttonsContainer, mcpIds);
-        }
-    }
-    if (!buttonsContainer.querySelector('.process-detail-btn')) {
         const progressDetailBtn = document.createElement('button');
         progressDetailBtn.className = 'mcp-detail-btn process-detail-btn';
         progressDetailBtn.innerHTML = '<span>' + (typeof window.t === 'function' ? window.t('chat.expandDetail') : '展开详情') + '</span>';
         progressDetailBtn.onclick = () => toggleProcessDetails(null, assistantMessageId);
-        buttonsContainer.appendChild(progressDetailBtn);
+        toolbar.appendChild(progressDetailBtn);
     }
 
-    // 创建详情容器，放在MCP按钮区域下方（统一结构）
     const detailsId = 'process-details-' + assistantMessageId;
     let detailsContainer = document.getElementById(detailsId);
+    const toolListEl = mcpSection.querySelector('.mcp-tool-list');
     
     if (!detailsContainer) {
         detailsContainer = document.createElement('div');
         detailsContainer.id = detailsId;
         detailsContainer.className = 'process-details-container';
-        // 确保容器在按钮容器之后
+        if (toolListEl) {
-        if (buttonsContainer.nextSibling) {
+            toolListEl.after(detailsContainer);
-            mcpSection.insertBefore(detailsContainer, buttonsContainer.nextSibling);
         } else {
             mcpSection.appendChild(detailsContainer);
         }
     }
     
-    // 设置详情内容（如果有错误，默认折叠；否则默认折叠）
     detailsContainer.innerHTML = `
         <div class="process-details-content">
             ${hasContent ? `<div class="progress-timeline" id="${detailsId}-timeline">${timelineHTML}</div>` : '<div class="progress-timeline-empty">' + (typeof window.t === 'function' ? window.t('chat.noProcessDetail') : '暂无过程详情（可能执行过快或未触发详细事件）') + '</div>'}
         </div>
     `;
     
-    // 确保初始状态是折叠的（默认折叠，特别是错误时）
     if (hasContent) {
         const timeline = document.getElementById(detailsId + '-timeline');
         if (timeline) {
-            // 如果有错误，确保折叠；否则也默认折叠
             timeline.classList.remove('expanded');
         }
         
@@ -1363,10 +1331,47 @@ function integrateProgressToMCPSection(progressId, assistantMessageId, mcpExecut
         });
     }
     
-    // 移除原来的进度消息（详情已快照到助手消息下的 process-details）
     removeMessage(progressId);
 }
 
+const PROCESS_DETAILS_PAGE_SIZE = 100;
+
+/**
+ * 分页加载过程详情并增量渲染，避免数百轮迭代一次性阻塞主线程。
+ */
+async function loadProcessDetailsPaginated(assistantMessageId, backendMessageId) {
+    if (!assistantMessageId || !backendMessageId || typeof apiFetch !== 'function' || typeof renderProcessDetails !== 'function') {
+        return;
+    }
+    const PAGE = PROCESS_DETAILS_PAGE_SIZE;
+    let offset = 0;
+    let isFirst = true;
+    while (true) {
+        const res = await apiFetch(
+            '/api/messages/' + encodeURIComponent(String(backendMessageId)) +
+            '/process-details?limit=' + PAGE + '&offset=' + offset
+        );
+        const j = await res.json().catch(() => ({}));
+        if (!res.ok) {
+            throw new Error((j && j.error) ? j.error : String(res.status));
+        }
+        const details = (j && Array.isArray(j.processDetails)) ? j.processDetails : [];
+        const hasMore = !!(j && j.hasMore);
+        renderProcessDetails(assistantMessageId, details, {
+            append: !isFirst,
+            markLoaded: !hasMore
+        });
+        if (!hasMore || details.length === 0) {
+            break;
+        }
+        offset += details.length;
+        isFirst = false;
+        await new Promise((resolve) => requestAnimationFrame(resolve));
+    }
+}
+
+window.loadProcessDetailsPaginated = loadProcessDetailsPaginated;
+
 // 切换过程详情显示
 function toggleProcessDetails(progressId, assistantMessageId) {
     const detailsId = 'process-details-' + assistantMessageId;
@@ -1383,26 +1388,17 @@ function toggleProcessDetails(progressId, assistantMessageId) {
                 // 正在加载中，避免重复请求
             } else {
                 detailsContainer.dataset.loading = '1';
-                // 先展开容器，显示加载态
                 const timeline = detailsContainer.querySelector('.progress-timeline');
                 if (timeline) {
                     timeline.innerHTML = '<div class="progress-timeline-empty">' + ((typeof window.t === 'function') ? window.t('common.loading') : '加载中…') + '</div>';
                 }
-                apiFetch(`/api/messages/${encodeURIComponent(String(backendMessageId))}/process-details`)
+                loadProcessDetailsPaginated(assistantMessageId, backendMessageId)
-                    .then(async (res) => {
-                        const j = await res.json().catch(() => ({}));
-                        if (!res.ok) throw new Error((j && j.error) ? j.error : res.status);
-                        const details = (j && Array.isArray(j.processDetails)) ? j.processDetails : [];
-                        // 重新渲染详情（renderProcessDetails 会清掉 lazy 标记并写入 loaded）
-                        renderProcessDetails(assistantMessageId, details);
-                    })
                     .catch((e) => {
                         console.error('加载过程详情失败:', e);
                         const tl = detailsContainer.querySelector('.progress-timeline');
                         if (tl) {
                             tl.innerHTML = '<div class="progress-timeline-empty">' + ((typeof window.t === 'function') ? window.t('chat.noProcessDetail') : '暂无过程详情（加载失败）') + '</div>';
                         }
-                        // 失败时保留 lazy 状态，允许用户重试
                         detailsContainer.dataset.lazyNotLoaded = '1';
                         detailsContainer.dataset.loaded = '0';
                     })
@@ -1944,6 +1940,7 @@ function handleStreamEvent(event, progressElement, progressId,
                 message: event.message || '',
                 data: d
             });
+            finalizeOutstandingToolCallsForProgress(progressId, 'failed');
             break;
         }
 
@@ -2755,6 +2752,9 @@ async function restoreHitlInlineForConversation(conversationId) {
             if (detailsContainer.dataset.lazyNotLoaded === '1' && detailsContainer.dataset.loaded !== '1') {
                 try {
                     detailsContainer.dataset.loading = '1';
+                    if (typeof loadProcessDetailsPaginated === 'function') {
+                        await loadProcessDetailsPaginated(clientMsgId, backendMsgId);
+                    } else {
                         const res = await apiFetch('/api/messages/' + encodeURIComponent(backendMsgId) + '/process-details');
                         const j = await res.json().catch(function () { return {}; });
                         if (!res.ok) throw new Error((j && j.error) ? j.error : String(res.status));
@@ -2762,6 +2762,7 @@ async function restoreHitlInlineForConversation(conversationId) {
                         if (typeof renderProcessDetails === 'function') {
                             renderProcessDetails(clientMsgId, details);
                         }
+                    }
                 } catch (e) {
                     console.error('加载过程详情失败（HITL 恢复）:', e);
                 } finally {
@@ -3531,6 +3532,7 @@ const monitorState = {
     timelineRange: null,
     timelineError: null,
     lastFetchedAt: null,
+    retentionDays: 0,
     pagination: {
         page: 1,
         pageSize: (() => {
@@ -3543,6 +3545,33 @@ const monitorState = {
     }
 };
 
+let monitorPollTimer = null;
+const MONITOR_POLL_INTERVAL_MS = 3000;
+
+function startMonitorPoll() {
+    stopMonitorPoll();
+    monitorPollTimer = setInterval(function () {
+        const page = document.getElementById('page-mcp-monitor');
+        if (!page || !page.classList.contains('active')) {
+            stopMonitorPoll();
+            return;
+        }
+        if (document.hidden) {
+            return;
+        }
+        if (typeof refreshMonitorPanel === 'function') {
+            refreshMonitorPanel().catch(function () { /* ignore */ });
+        }
+    }, MONITOR_POLL_INTERVAL_MS);
+}
+
+function stopMonitorPoll() {
+    if (monitorPollTimer) {
+        clearInterval(monitorPollTimer);
+        monitorPollTimer = null;
+    }
+}
+
 function openMonitorPanel() {
     // 切换到MCP监控页面
     if (typeof switchPage === 'function') {
@@ -3626,6 +3655,7 @@ async function refreshMonitorPanel(page = null) {
         monitorState.timeline = timeline;
         monitorState.timelineError = timelineError;
         monitorState.lastFetchedAt = new Date();
+        monitorState.retentionDays = typeof result.retention_days === 'number' ? result.retention_days : 0;
         
         // 更新分页信息
         if (result.total !== undefined) {
@@ -3709,6 +3739,7 @@ async function refreshMonitorPanelWithFilter(statusFilter = 'all', toolFilter =
         monitorState.timeline = timeline;
         monitorState.timelineError = timelineError;
         monitorState.lastFetchedAt = new Date();
+        monitorState.retentionDays = typeof result.retention_days === 'number' ? result.retention_days : 0;
         
         // 更新分页信息
         if (result.total !== undefined) {
@@ -4526,15 +4557,20 @@ function renderMcpStatsStackedBar(success, failed) {
     </div>`;
 }
 
-function updateMonitorStatsSubtitle(lastFetchedAt, toolCount) {
+function updateMonitorStatsSubtitle(lastFetchedAt, toolCount, retentionDays) {
     const subtitle = document.getElementById('monitor-stats-subtitle');
     if (!subtitle) return;
     const locale = (typeof window.__locale === 'string' && window.__locale.startsWith('zh')) ? 'zh-CN' : 'en-US';
     const timeText = lastFetchedAt
         ? (lastFetchedAt.toLocaleString ? lastFetchedAt.toLocaleString(locale) : String(lastFetchedAt))
         : '—';
-    const text = mcpMonitorT('statsSubtitle', { time: timeText, count: toolCount })
+    let text = mcpMonitorT('statsSubtitle', { time: timeText, count: toolCount })
         || monitorFallback(`最后刷新 ${timeText} · 共 ${toolCount} 个工具`, `Refreshed ${timeText} · ${toolCount} tools`);
+    if (typeof retentionDays === 'number' && retentionDays > 0) {
+        const hint = mcpMonitorT('retentionHint', { days: retentionDays })
+            || monitorFallback(`执行记录保留 ${retentionDays} 天，超期自动清理`, `Execution records are kept for ${retentionDays} days, then purged automatically.`);
+        text += ' · ' + hint;
+    }
     subtitle.textContent = text;
     subtitle.hidden = false;
 }
@@ -4959,7 +4995,7 @@ function renderMonitorStats(statsMap = {}, lastFetchedAt = null) {
     } else if (toolFilterEl) {
         toolFilterEl.classList.remove('is-filter-active');
     }
-    updateMonitorStatsSubtitle(lastFetchedAt, entries.length);
+    updateMonitorStatsSubtitle(lastFetchedAt, entries.length, monitorState.retentionDays);
 }
 
 function renderMonitorExecutions(executions = [], statusFilter = 'all') {
@@ -5459,6 +5495,22 @@ function refreshProgressAndTimelineI18n() {
         const expanded = timeline && timeline.classList.contains('expanded');
         span.textContent = expanded ? _t('tasks.collapseDetail') : _t('chat.expandDetail');
     });
+
+    document.querySelectorAll('#chat-messages .message.assistant').forEach(function (msgEl) {
+        if (typeof window.syncMcpToolsToggleButton === 'function') {
+            window.syncMcpToolsToggleButton(msgEl);
+        }
+    });
+
+    const copyLabel = _t('common.copy');
+    const copyTitle = _t('chat.copyMessageTitle');
+    document.querySelectorAll('#chat-messages .message-copy-btn').forEach(function (btn) {
+        if (btn.dataset.copySuccessActive === '1') return;
+        const span = btn.querySelector('span');
+        if (span) span.textContent = copyLabel;
+        btn.title = copyTitle;
+        btn.setAttribute('aria-label', copyTitle);
+    });
 }
 
 document.addEventListener('languagechange', function () {

web/static/js/router.js

@@ -356,6 +356,9 @@ async function initPage(pageId) {
             if (typeof refreshMonitorPanel === 'function') {
                 refreshMonitorPanel();
             }
+            if (typeof startMonitorPoll === 'function') {
+                startMonitorPoll();
+            }
             break;
         case 'mcp-management':
             // 初始化MCP管理

web/static/js/tasks.js

@@ -990,6 +990,7 @@ async function createBatchQueue() {
     const roleSelect = document.getElementById('batch-queue-role');
     const projectSelect = document.getElementById('batch-queue-project-id');
     const agentModeSelect = document.getElementById('batch-queue-agent-mode');
+    const concurrencyInput = document.getElementById('batch-queue-concurrency');
     const scheduleModeSelect = document.getElementById('batch-queue-schedule-mode');
     const cronExprInput = document.getElementById('batch-queue-cron-expr');
     const executeNowCheckbox = document.getElementById('batch-queue-execute-now');
@@ -1019,6 +1020,9 @@ async function createBatchQueue() {
     const scheduleMode = scheduleModeSelect ? (scheduleModeSelect.value === 'cron' ? 'cron' : 'manual') : 'manual';
     const cronExpr = cronExprInput ? cronExprInput.value.trim() : '';
     const executeNow = executeNowCheckbox ? !!executeNowCheckbox.checked : false;
+    let concurrency = concurrencyInput ? parseInt(concurrencyInput.value, 10) : 1;
+    if (!Number.isFinite(concurrency) || concurrency < 1) concurrency = 1;
+    if (concurrency > 8) concurrency = 8;
     if (scheduleMode === 'cron' && !cronExpr) {
         alert(_t('batchImportModal.cronExprRequired'));
         return;
@@ -1043,6 +1047,7 @@ async function createBatchQueue() {
                 cronExpr,
                 executeNow,
                 projectId,
+                concurrency,
             }),
         });
         
@@ -1489,6 +1494,7 @@ async function showBatchQueueDetail(queueId) {
                 <div class="bq-kv"><span class="bq-kv__k">${escapeHtml(_t('batchQueueDetailModal.role'))}</span><span class="bq-kv__v" id="bq-role-val">${allowSubtaskMutation ? `<span class="bq-inline-editable" onclick="startInlineEditRole()" title="${escapeHtml(_t('common.edit'))}">${roleLineVal}</span>` : roleLineVal}</span></div>
                 <div class="bq-kv"><span class="bq-kv__k">${escapeHtml(_t('batchImportModal.agentMode'))}</span><span class="bq-kv__v" id="bq-agentmode-val">${allowSubtaskMutation ? `<span class="bq-inline-editable" onclick="startInlineEditAgentMode()" title="${escapeHtml(_t('common.edit'))}">${escapeHtml(agentModeText)}</span>` : escapeHtml(agentModeText)}</span></div>
                 <div class="bq-kv"><span class="bq-kv__k">${escapeHtml(_t('batchImportModal.scheduleMode'))}</span><span class="bq-kv__v" id="bq-schedule-val">${allowSubtaskMutation ? `<span class="bq-inline-editable" onclick="startInlineEditSchedule()" title="${escapeHtml(_t('common.edit'))}">${scheduleDetail}</span>` : scheduleDetail}</span></div>
+                <div class="bq-kv"><span class="bq-kv__k">${escapeHtml(_t('batchQueueDetailModal.concurrency'))}</span><span class="bq-kv__v" id="bq-concurrency-val">${allowSubtaskMutation ? `<span class="bq-inline-editable" onclick="startInlineEditConcurrency()" title="${escapeHtml(_t('common.edit'))}">${escapeHtml(String(queue.concurrency && queue.concurrency > 0 ? queue.concurrency : 1))}</span>` : escapeHtml(String(queue.concurrency && queue.concurrency > 0 ? queue.concurrency : 1))}</span></div>
                 <div class="bq-kv"><span class="bq-kv__k">${escapeHtml(_t('batchQueueDetailModal.taskTotal'))}</span><span class="bq-kv__v">${queue.tasks.length}</span></div>
                 ${queue.scheduleMode === 'cron' ? `<div class="bq-kv bq-kv--block"><span class="bq-kv__k">${escapeHtml(_t('batchQueueDetailModal.scheduleCronAuto'))}</span><span class="bq-kv__v bq-kv__v--control"><label class="bq-cron-toggle"><input type="checkbox" ${queue.scheduleEnabled !== false ? 'checked' : ''} onchange="updateBatchQueueScheduleEnabled(this.checked)" /><span class="bq-cron-toggle__hint">${escapeHtml(_t('batchQueueDetailModal.scheduleCronAutoHint'))}</span></label></span></div>` : ''}
             </section>
@@ -2287,6 +2293,75 @@ async function saveInlineAgentMode() {
     }
 }
 
+function normalizeBatchQueueConcurrencyInput(raw) {
+    let n = parseInt(raw, 10);
+    if (!Number.isFinite(n) || n < 1) n = 1;
+    if (n > 8) n = 8;
+    return n;
+}
+
+// --- 内联编辑：并发数 ---
+function startInlineEditConcurrency() {
+    const container = document.getElementById('bq-concurrency-val');
+    if (!container) return;
+    const queueId = batchQueuesState.currentQueueId;
+    if (!queueId) return;
+    apiFetch(`/api/batch-tasks/${queueId}`).then(r => r.json()).then(detail => {
+        const queue = detail.queue || {};
+        const current = normalizeBatchQueueConcurrencyInput(queue.concurrency || 1);
+        container.innerHTML = `<span class="bq-inline-edit-controls">
+            <input type="number" id="bq-edit-concurrency" min="1" max="8" value="${current}" style="width:72px;" />
+        </span>`;
+        const inp = document.getElementById('bq-edit-concurrency');
+        if (!inp) return;
+        inp.focus();
+        inp.select();
+        let cancelled = false;
+        inp.addEventListener('keydown', (e) => {
+            if (e.key === 'Enter') { e.preventDefault(); inp.blur(); }
+            if (e.key === 'Escape') { cancelled = true; cancelAllInlineEdits(); }
+        });
+        inp.addEventListener('blur', () => {
+            if (!cancelled) saveInlineConcurrency();
+        });
+    });
+}
+
+async function saveInlineConcurrency() {
+    if (_bqInlineSaving) return;
+    _bqInlineSaving = true;
+    const queueId = batchQueuesState.currentQueueId;
+    if (!queueId) { _bqInlineSaving = false; return; }
+    const inp = document.getElementById('bq-edit-concurrency');
+    const concurrency = normalizeBatchQueueConcurrencyInput(inp ? inp.value : 1);
+    try {
+        const detailResp = await apiFetch(`/api/batch-tasks/${queueId}`);
+        const detail = await detailResp.json();
+        const q = detail.queue || {};
+        const response = await apiFetch(`/api/batch-tasks/${queueId}/metadata`, {
+            method: 'PUT',
+            headers: { 'Content-Type': 'application/json' },
+            body: JSON.stringify({
+                title: q.title || '',
+                role: q.role || '',
+                agentMode: q.agentMode || 'eino_single',
+                concurrency,
+            }),
+        });
+        if (!response.ok) {
+            const result = await response.json().catch(() => ({}));
+            throw new Error(result.error || _t('tasks.updateTaskFailed'));
+        }
+        _bqInlineSaving = false;
+        showBatchQueueDetail(queueId);
+        refreshBatchQueues();
+    } catch (e) {
+        _bqInlineSaving = false;
+        console.error(e);
+        alert(e.message);
+    }
+}
+
 // --- 单条执行 ---
 async function runSingleBatchTask(queueId, taskId) {
     if (!queueId || !taskId) return;
@@ -2441,6 +2516,8 @@ window.startInlineEditRole = startInlineEditRole;
 window.saveInlineRole = saveInlineRole;
 window.startInlineEditAgentMode = startInlineEditAgentMode;
 window.saveInlineAgentMode = saveInlineAgentMode;
+window.startInlineEditConcurrency = startInlineEditConcurrency;
+window.saveInlineConcurrency = saveInlineConcurrency;
 window.runSingleBatchTask = runSingleBatchTask;
 window.startInlineEditSchedule = startInlineEditSchedule;
 window.toggleInlineScheduleCron = toggleInlineScheduleCron;

web/templates/index.html

@@ -3777,7 +3777,9 @@
             <div class="modal-body batch-manage-body">
                 <div class="batch-conversations-table">
                     <div class="batch-table-header">
-                        <div class="batch-table-col-checkbox"></div>
+                        <div class="batch-table-col-checkbox">
+                            <input type="checkbox" id="batch-select-all" onchange="toggleSelectAllBatch()" data-i18n="batchManageModal.selectAll" data-i18n-attr="title" title="全选" />
+                        </div>
                         <div class="batch-table-col-name" data-i18n="batchManageModal.conversationName">对话名称</div>
                         <div class="batch-table-col-time" data-i18n="batchManageModal.lastTime">最近一次对话时间</div>
                         <div class="batch-table-col-action" data-i18n="batchManageModal.action">操作</div>
@@ -3786,10 +3788,6 @@
                 </div>
             </div>
             <div class="modal-footer batch-manage-footer">
-                <label class="select-all-checkbox">
-                    <input type="checkbox" id="batch-select-all" onchange="toggleSelectAllBatch()" />
-                    <span data-i18n="batchManageModal.selectAll">全选</span>
-                </label>
                 <div class="batch-footer-actions">
                     <button class="btn-secondary" onclick="closeBatchManageModal()" data-i18n="common.cancel">取消</button>
                     <button class="btn-primary" onclick="deleteSelectedConversations()" data-i18n="batchManageModal.deleteSelected">删除所选</button>
@@ -4012,6 +4010,11 @@
                     </select>
                     <div class="form-hint" style="margin-top: 4px;" data-i18n="batchImportModal.agentModeHint">与对话页一致：Eino 单代理（ADK），或 Deep / Plan-Execute / Supervisor（后三种需已启用多代理）。</div>
                 </div>
+                <div class="form-group">
+                    <label for="batch-queue-concurrency" data-i18n="batchImportModal.concurrency">并发数</label>
+                    <input type="number" id="batch-queue-concurrency" min="1" max="8" value="1" style="width: 100%; padding: 8px; border: 1px solid #ddd; border-radius: 4px; font-size: 0.875rem;" />
+                    <div class="form-hint" style="margin-top: 4px;" data-i18n="batchImportModal.concurrencyHint">同时执行的子任务数量（1-8）。默认 1 为串行；含扫描类工具时建议 1-2。</div>
+                </div>
                 <div class="form-group">
                     <label for="batch-queue-schedule-mode" data-i18n="batchImportModal.scheduleMode">调度方式</label>
                     <select id="batch-queue-schedule-mode" onchange="handleBatchScheduleModeChange()" style="width: 100%; padding: 8px; border: 1px solid #ddd; border-radius: 4px; font-size: 0.875rem;">