도구

One plugin bundling skills and subagents for deep research, fixing GitHub issues and reviewing PRs, Codex handoff, Gemini or GPT image generation, translation and YouTube transcript extraction, plus settings.json templates for various providers (manual setup). Some skills need API keys.

Lets agents drive WinDbg's CDB/KD to analyze crash dumps, debug user-mode processes remotely and debug kernels, and triage many dumps in a batch. Windows only; requires WinDbg or Debugging Tools for Windows.

Microsoft's official Azure skill set: deployment prep and validation, deployment, production diagnostics, AKS, quota lookup, compliance audits, cloud migration and Foundry; the plugin version also connects the Azure MCP Server. Needs an Azure account and az login.

Microsoft's official Azure plugin: bundles 20+ skills such as azure-prepare, azure-deploy, azure-diagnostics and azure-cost, and connects the Azure MCP and Foundry MCP to query resources, estimate costs, deploy and troubleshoot. Needs an Azure subscription and az login.

Docker's official MCP gateway (the docker mcp CLI plugin): runs servers from the Docker MCP Catalog in isolated containers, manages secrets, OAuth and tool toggles centrally, and gives multiple clients one endpoint. Built into Docker Desktop 4.59+; Docker CE and other setups must build the plugin.

Figma's official CLI that links Figma design-system components to real components in your code, so Dev Mode and the Figma MCP show production snippets instead of generated examples. Needs Node.js 18+, a Figma Organization or Enterprise plan, a full Design or Dev Mode seat and an access token.

MiniMax's official MCP: text-to-speech, voice cloning and voice design, text-to-image and video generation. Needs a MiniMax API key; the key and API host must belong to the same region (global or mainland China). Pay as you go.

Uses unified rule files in a .rulesync directory as the source to generate rules, MCP, commands, subagents, skills and hooks configs for dozens of AI coding tools including Claude Code, Cursor, Copilot and Codex. Can also import existing configs or convert directly between tools.

Web search through DuckDuckGo plus fetching and parsing page content, with built-in rate limiting and LLM-friendly result formatting. No API key needed; runs with a single uvx command.

Brings the Brave Search API to agents: web, news, image, video and local business and place search, plus LLM-ready page context and AI summaries. Maintained by Brave; needs an API key.

Uses the Qdrant vector database as a semantic memory layer for agents: store information and retrieve it by meaning, connecting to a remote instance or a local path, with FastEmbed generating embeddings by default. Maintained by Qdrant.

Condenses verbose xcodebuild and swift test output into concise build errors, warnings and test results, so agents read less log noise. Can generate JUnit reports and works with CI such as GitHub Actions; runs on macOS and Linux.

Batch-normalizes audio and video loudness to standards like EBU R128, with presets for podcasts, music and streaming video; handy for evening out course recordings, podcasts and videos before publishing. Needs FFmpeg and Python 3.10+.

Connects LLMs to ROS and ROS 2 robots: publish and subscribe to topics, call services and actions, set parameters, and read sensor and robot state, without changing robot code. The robot side needs rosbridge running.

A plugin marketplace with dozens of plugins: an enhanced skill-creator, scrapers for Bilibili, Douban and Feishu docs, China A-share news and US stock fundamentals, WeCom notifications, deep research and fact-checking, with docs in English and Chinese.

Sets up and audits i18n for React, Next.js and Vue apps: replaces hard-coded strings with message keys, checks that English and Chinese keys are complete and consistent, and verifies plurals and date and number formats, with a key audit script. Needs a local codebase and Python.

Audio plugin suite: transcription with local Qwen3-ASR or Whisper (MLX on Apple Silicon), speaker diarization, transcript correction, structured meeting notes and StepFun TTS. StepFun ASR and TTS need an API key.

Lets Claude Code watch videos: extracts frames with ffmpeg and transcribes timestamped audio from local files or YouTube links, to review footage, summarize lectures or find on-screen text. Needs Node.js 20+ and ffmpeg (yt-dlp for YouTube); transcribe with local Whisper or a Gemini/OpenAI key.

Stem separation CLI built on UVR models: splits audio into vocals and accompaniment in one command, isolates drums, bass and other stems, or removes noise and reverb; handy for stripping background music or extracting clean vocals. Models download on first use; needs Python and FFmpeg.

Adds voice conversations to Claude Code and other agents: speak, and it detects silence, transcribes, and reads replies aloud. Runs offline with local Whisper and Kokoro, or uses OpenAI's speech APIs; needs a microphone and speakers.

Opens HTML, Markdown or localhost pages in the browser via npx human-review so you can edit text directly and leave anchored comments, then sends every edit and comment back to the agent in one batch to apply to the source.

Gives agents controlled access to MySQL: list databases and tables, inspect schemas and sample data, and run SQL, with multi-database mode, SSL and SSH tunnels. Configure your own database user, ideally with least privilege.

Generates book covers, movie posters, article covers and social images from one sentence, picking colors and composition from 30+ poster masters' styles to suit the subject. Rendering directly needs Python and a PipeLLM image API key; or just generate prompts for another image tool.

Local transcription that's compatible with the OpenAI Whisper CLI but faster and lighter: outputs SRT/VTT subtitles or text and can transcribe a live microphone. Needs Python; models download on first run; speaker labels need pyannote.audio and a Hugging Face token.