Merge branch 'main' into add-mirrord

README.md

@@ -143,6 +143,8 @@ Lightweight command-line tools for AI-assisted commits, shell translation, and w
 - [resume-cli](https://github.com/inevolin/resume-cli) — CLI that aggregates recent sessions from Claude Code, Codex, and GitHub Copilot in one place. Pick a session and resume it in any of the three tools.
 - [codesight](https://github.com/Houseofmvps/codesight) — CLI token optimizer and AI context generator. Scans codebases to extract routes, schema, components, and dependencies for Claude Code, Cursor, Copilot, Codex, and Windsurf. 9x–13x token reduction, built-in MCP server, zero runtime dependencies. `npx codesight`
 - [CLIRank](https://clirank.dev) — API directory scoring 387 APIs on agent-friendliness across 11 signals. Available as an MCP server (`clirank-mcp-server`) and REST API.
+- [prpack](https://github.com/Lucas2944/prpack) — Zero-dep CLI that packs a pull request (diff + commits + full post-change file contents) into one markdown file optimized for LLM code review. Drop the file into Claude / Cursor / your model. MIT.
+- [Marmot](https://marmot.sh) — Shell-native CLI that gives agents one command shape for AI, web search, scraping, and data enrichment across many providers. Designed for Claude Code, Codex, OpenCode, and similar harnesses; composes via shell pipes.
 
 ---
 
@@ -266,6 +268,7 @@ Native desktop apps for AI-powered coding, terminal enhancement, and agent orche
 - [Parallel Code](https://github.com/johannesjo/parallel-code) - Desktop app that runs multiple AI coding agents (Claude Code, Codex CLI, Gemini CLI) in parallel with automatic git worktree isolation, a unified GUI, and remote monitoring.
 - [PATAPIM](https://patapim.ai) — Terminal IDE for AI coding agents with a 9-terminal grid, AI state detection, built-in Whisper voice dictation, LAN remote access, and embedded MCP browser. Built with Electron.
 - [IM.codes](https://github.com/im4codes/imcodes) — Mobile/web control layer for Claude Code, Codex, Gemini CLI, and other terminal-based coding agents, built for away-from-desk continuation with terminal access, file browsing, git views, localhost preview, notifications, and multi-agent workflows.
+- [OpenQuack](https://github.com/larryxiao/openquack) - Privacy-first local voice dictation menu bar app for macOS. Pairs with Claude Code, Cursor, Codex, and Aider to dictate long contextual prompts; transcribes via WhisperKit on Apple Silicon, pastes at the cursor, ~8 MB, MIT.
 
 ---
 
@@ -306,9 +309,11 @@ Integrations that automatically review pull requests and suggest code fixes:
 - [MiniMax Code Review](https://github.com/tarmojussila/minimax-code-review) - AI-powered GitHub Pull Request code review using MiniMax models.
 - [claude-pr-reviewer](https://github.com/indoor47/claude-pr-reviewer) - GitHub Action + CLI that uses Claude AI to automatically review pull requests, post inline comments, and report token costs. Supports strictness levels (lenient/balanced/strict). Zero dependencies, Python 3.8+.
 - [Revieko](https://synqra.tech/revieko/) — Architecture drift radar for PRs: structural risk scoring and drift hotspots.
+- [crit](https://crit.md) - Human-in-the-loop review for AI coding agents. Annotate plans, code, or any file with inline comments in a local browser UI; exports structured JSON your agent reads back. Share reviews via URL.
 - [CodeHawk](https://codehawk.crossgen-ai.com) — GitHub App that installs in seconds and reviews pull requests automatically, posting inline comments on bugs, security vulnerabilities, and logic errors.
 - [Grit](https://app.grit.io) — GitHub-integrated agent for automating maintenance tasks and other development work.
 - [PR Triage](https://pr-triage-web.vercel.app) — Open source PR evaluation tool that scores pull requests on six quality dimensions with diff evidence. BYOK, MIT licensed. [Source](https://github.com/Elifterminal/pr-triage-web)
+- [prpack-action](https://github.com/Lucas2944/prpack-action) — GitHub Action that runs prpack on every PR, uploads the packed markdown as an artifact, and posts a summary comment. MIT.
 
 ### CI/CD & Testing Automation
 
@@ -364,6 +369,7 @@ Secure isolated environments for running AI coding agents with controlled access
 - [Open Agent](https://github.com/Th0rgal/openagent) — Self-hosted control plane for Claude Code with isolated container workspaces and real-time mission streaming.
 - [FlyDex](https://flydex.net) — Browser-first remote control plane for local Codex sessions with QR pairing, approvals, and machine continuity.
 - [mirrord](https://github.com/metalbear-co/mirrord) — Per-agent isolation inside a shared Kubernetes cluster: traffic filters, DB branches, and Kafka queue splits via the mirrord Operator. Six [Claude Code skills](https://github.com/metalbear-co/skills) cover quickstart, config, operator setup, CI, DB branching, and Kafka splitting; install via `/plugin marketplace add metalbear-co/skills`.
+- [AgentTier](https://github.com/agenttier/agenttier) — Open-source, Kubernetes-native sandbox runtime for AI coding agents (Claude Code, LangGraph, OpenHands). Each sandbox is a Pod + PVC + default-deny NetworkPolicy with optional gVisor isolation; runs in interactive `mode: code` (browser terminal) or `mode: agent` (REST `/configure` + SSE-streaming `/invoke`). Apache-2.0.
 
 ### Configuration & Context Management
 
@@ -372,8 +378,10 @@ Tools that manage and sync AI agent configurations, rules, and context across ed
 - [Context7](https://context7.com/) — Documentation platform that provides up-to-date, version-specific documentation and code examples for any library directly into Cursor, Claude Code, Windsurf, and other AI coding tools.
 - [cursor-doctor](https://github.com/nedcodes-ok/cursor-doctor) — Health scanner and diagnostic tool for Cursor AI IDE configuration (.cursor/rules/ files). Scans for broken YAML frontmatter, rule conflicts, redundancies, missing alwaysApply, token budget issues, and stack coverage gaps. Available as CLI (`npx cursor-doctor scan`) and VS Code extension.
 - [ctxlint](https://github.com/YawLabs/ctxlint) — Open-source linter for AI context files (CLAUDE.md, .cursorrules, copilot-instructions.md) that catches stale paths, wrong commands, and token waste by validating against the real codebase.
+- [Entroly](https://github.com/juyterman1000/entroly) - Open-source context optimization engine that cuts AI token costs by 70-95%. Uses submodular knapsack selection and PRISM reinforcement learning to provide the exact context needed to 65+ supported AI coding agents. Features a built-in MCP server, semantic caching, and SimHash deduplication. Apache-2.0.
 - [lean-ctx](https://github.com/yvgude/lean-ctx) — Open-source context runtime for AI coding agents. MCP server (optional shell hook) that compresses file reads, shell output, and codebase search to reduce token usage, often 60–99% on supported workflows. 46 tools, session caching, AST-aware compression for 18 languages, 90+ shell patterns. [Website](https://leanctx.com). Apache-2.0.
 - [Caliber](https://github.com/rely-ai-org/caliber) — Open-source CLI that scans your codebase and generates AI agent configs (CLAUDE.md, .cursorrules, skills, MCPs) for Claude Code, Cursor, and Codex. Scores your setup 0-100 and recommends MCP servers.
+- [claude-overlay](https://github.com/mzmmoazam/claude-overlay) — CLI for managing Claude Code project configs across custom providers (Databricks, Bedrock, OpenRouter, LiteLLM, Cloudflare). Handles overlay merge/remove, web search MCP setup, and multi-provider switching.
 - [vsync](https://github.com/nicepkg/vsync) — CLI tool that syncs Skills, MCP servers, Agents & Commands across Claude Code, Cursor, OpenCode, and Codex with automatic format conversion (JSON <-> TOML <-> JSONC).
 - [rule-porter](https://github.com/nedcodes-ok/rule-porter) — Zero-dependency CLI that converts AI IDE rule files between Cursor (.mdc), CLAUDE.md, AGENTS.md, Copilot, and Windsurf. Bidirectional with lossy-conversion warnings.
 - [LynxPrompt](https://github.com/GeiserX/LynxPrompt) — Self-hostable platform for managing AI IDE configuration files. Generates, syncs, and shares configs (.cursorrules, CLAUDE.md, copilot-instructions.md, etc.) across 30+ AI coding assistants via web UI, REST API, CLI, and federated blueprint marketplace.
@@ -392,6 +400,8 @@ Tools that manage and sync AI agent configurations, rules, and context across ed
 - [Agentify](https://github.com/koriyoshi2041/agentify) — CLI tool that transforms any OpenAPI spec into 9 agent interface formats (MCP server, AGENTS.md, CLAUDE.md, .cursorrules, Skills, llms.txt, GEMINI.md, A2A Card, CLI) with a single command. Tiered generation strategies for small to large APIs.
 - [skill-optimizer](https://github.com/fastxyz/skill-optimizer) — CLI that benchmarks SDK, CLI, and MCP guidance docs (SKILL.md) against multiple LLMs and runs an iterative optimizer to rewrite them until every configured model meets a score floor.
 - [KubeStellar Console kc-agent](https://github.com/kubestellar/console) — MCP server that bridges AI coding agents (Claude Code, Copilot, Codex) to multi-cluster Kubernetes APIs. Enables natural language queries across clusters, workload placement, policy enforcement, and real-time observability. CNCF Sandbox project.
+- [claude-code-pro-pack](https://github.com/sisyphusse1-ops/claude-code-pro-pack) — Drop-in 12-rule `CLAUDE.md` + `AGENTS.md` baseline that closes common agent-orchestration failures (token spirals, silent partial failures, two-pattern pollution). Includes PRD generator prompt, browser-skill-graduation workflow, and 5 example skills. ~700 tokens total, MIT.
+- [cc-audit](https://github.com/sisyphusse1-ops/cc-audit) — Single-file Python linter that scores any `CLAUDE.md` / `AGENTS.md` against a 12-rule baseline. Flags leaked secrets (GitHub PATs, AWS keys, PayPal links), the 200-line compliance cliff, and missing project-specifics sections. Zero dependencies, JSON output for CI, MIT.
 
 ### Usage Analytics & Cost Tracking
 
@@ -400,13 +410,16 @@ Tools for monitoring token usage and API costs across AI providers:
 - [Tokscale](https://github.com/junhoyeo/tokscale) — CLI tool for tracking token usage from AI coding agents (OpenCode, Claude Code, OpenClaw, Codex, Gemini CLI, Cursor IDE, AmpCode, Factory Droid) with a global leaderboard and 2D/3D contribution graphs.
 - [BurnRate](https://getburnrate.io) - Local-first AI coding cost analytics. Tracks Claude Code, Cursor, Codex, Copilot, Windsurf, Cline, and Aider. Cost breakdowns, 23 optimization rules, rate limit monitoring, provider comparison, and PDF reports.
 - [Code Insights](https://github.com/melagiri/code-insights) — Local-first CLI and dashboard for analyzing AI coding sessions from Claude Code, Cursor, Codex CLI, Copilot CLI, and VS Code Copilot Chat. SQLite-backed with terminal analytics, browser dashboard, and LLM-powered insights.
+- [agenttrace](https://github.com/luoyuctl/agenttrace) — Local-first TUI observability for AI coding agents. Tracks cost, tokens, tool failures, anomalies, health, and CI gates across Claude Code, Codex CLI, Gemini CLI, Aider, and Cursor exports.
 - [onWatch](https://github.com/onllm-dev/onwatch) — Open-source Go CLI that tracks AI API quota usage across 7 providers (Anthropic, OpenAI Codex, GitHub Copilot, Synthetic, Z.ai, MiniMax, Antigravity). Background daemon with Material Design 3 web dashboard, ~15MB binary, <50MB RAM, zero telemetry.
 - [claude-lens](https://github.com/Astro-Han/claude-lens) — Lightweight Bash + jq statusline for Claude Code with rate limit pace tracking. Compares burn rate to time remaining in 5h/7d windows, showing whether usage is ahead or behind sustainable pace. Single file, zero dependencies beyond jq.
 - [cc-tempo](https://github.com/O0000-code/cc-tempo) — Claude Code statusline that parses transcript JSONL for active work time (excludes idle and interactive-tool waits), SubAgent speedup ratio, /clear-resilient PID timer, code-churn sparkline with trend arrow, and multi-instance detection. Bash + TypeScript CLI, MIT.
 - [ai-coding-tools-pricing](https://github.com/lunacompsia-oss/ai-coding-tools-pricing) — Open JSON dataset of pricing for 8 AI coding tools (Copilot, Cursor, Claude Code, Windsurf, etc.). 30+ tiers, TypeScript types, JSON Schema. CC-BY-4.0. Updated monthly.
 - [CodeCosts](https://codecosts.pages.dev) — Interactive cost calculator and comparison tool for AI coding tools. Uses the ai-coding-tools-pricing dataset to help developers pick the right plan.
+- [WhereMyTokens](https://github.com/jeongwookie/WhereMyTokens) — Windows system tray app for monitoring Claude Code token usage in real time. Displays per-session token counts, costs, context window usage, and tool call breakdowns; fetches 5h/1w rate-limit data from the Anthropic API with optional statusLine bridge integration for live data without polling.
 - [aicost](https://github.com/dwylq/aicost) — Universal AI coding cost analyzer CLI. Scans Claude Code, Cursor, and GitHub Copilot usage with cache-aware pricing, HTML dashboard, and cost alerting. No API key required.
 - [CostGoat](https://costgoat.com) — Privacy-first menubar app for tracking AI agent quotas (Claude Code, Codex, Kimi, Z.ai) and LLM API costs (OpenAI, OpenRouter, Anthropic, ElevenLabs) in real-time. Also covers cloud spend and SaaS subscriptions.
+- [TokenWise](https://github.com/CodeShuX/tokenwise) — Measurement-driven model router for Claude Code. Auto-routes subtasks across Haiku/Sonnet/Opus based on task class, logs every routed task with verified $ saved to a local NDJSON, and includes an A/B test subcommand to validate cheaper tiers before trusting routing. MIT, zero telemetry, Anthropic-only.
 
 ---
 
@@ -455,3 +468,5 @@ Curated lists, comparison guides, and configuration templates for AI coding tool
 - [Havoptic](https://havoptic.com/) — Free, open-source timeline tracking releases from AI coding tools. Auto-updated daily. [Source](https://github.com/scotthavird/havoptic.com)
 - [AI Dev Toolkit](https://github.com/dohko04/awesome-ai-prompts-for-devs) — 198 production-ready AI engineering resources: prompt frameworks, multi-agent configs, MCP setups, RAG pipelines, and automation scripts for Claude Code, Cursor, GPT-5, Copilot, and more. Free samples included.
 - [AI Coding Guide (中文)](https://github.com/jnMetaCode/ai-coding-guide) — Chinese best practices for 8 AI coding tools (Claude Code, Cursor, Copilot, Windsurf, Gemini CLI, Kiro, Aider, Trae) with copy-paste config templates and multi-tool collaboration workflows.
+- [Claude Code Skills 中文精选集](https://claude-skills.bt199.com/) — Chinese curated directory of Claude Code Skills, agents, plugins, and workflow resources for developers using Claude Code.
+- [Awesome AI Startups — Coding & Developer Tools](https://github.com/nowork-studio/awesome-ai-startups#-coding--developer-tools) — 100+ indie-built AI coding assistants, dev tools, and code-generation products from bootstrapped, pre-seed, and angel-funded startups.