每日 GitHub + registry 抓取

中文 EN

来源方法

加载中…

找到能用的 MCP 和 agent skills。

按能力搜索，直接比较来源、安装方式、新鲜度和质量分。

只看 MCP servers 只看 agent skills 官方来源2,809 活跃条目

筛选点开筛选

仅官方仅可直接安装仅实测可连通包含实测无法连通的条目

重置

筛选

仅官方仅可直接安装仅实测可连通包含实测无法连通的条目重置

MCP仅源码GitHub MCP Searchai-codingclaude-code

mcclaude

Develop Minecraft plugins with Claude Code on a live server. Read console, run commands, edit Skript files, and eval code — all from your IDE. End-to-end encrypted bridge between Claude Code and your Minecraft server.

LightningReflex/mcclaudeMIT2026年9月19日

评分48

质量分如何计算？

由六个维度加权得出，满分 100。
来源可信度（25）：是否收录于官方或精选 registry。
可安装性（25）：可直接运行的命令 > 已声明传输方式 > 仅源码。
维护活跃度（20）：最近推送时间；已归档条目此项为 0。
采用度（20）：仓库 star 数，按对数曲线计分。
文档完整性（10）与描述质量（10）：是否有已解析的 manifest 或 SKILL.md frontmatter，以及是否为真实描述而非占位文本。
某个来源结构上无法提供的维度（registry 不提供 star 和许可证）不计入该条目的平均分，而不是记为 0 分，因此不同来源的条目仍可横向比较。
安全信号最多扣 20 分。每日抓取时重新计算。

仓库 Stars10

仓库更新2026年9月19日

git clone https://github.com/LightningReflex/mcclaude.git详情

运行此命令前，请先核对上游源码。

对比

MCP仅源码GitHub MCP Searchcommunitymcp

feishu-mcp-server

飞书 MCP Server — 让 Claude Code 以你的身份直接操作飞书文档、发送消息、调用任意飞书 API。支持 OAuth 用户授权。Feishu/Lark MCP Server — Let Claude Code read & write your Feishu docs, send messages, and call any Feishu API, using your identity via OAuth.

zjdx10/feishu-mcp-serverMIT2026年8月26日

评分44

质量分如何计算？

MCP仅源码GitHub MCP Searchai-memorychatgpt

mcp-memory-server

Long-term memory for AI coding agents. Works with Claude Code, ChatGPT, Cursor, Gemini CLI, VS Code Copilot, Windsurf, Claude Desktop, Codex CLI. Deploy once to Cloudflare Workers (free), use everywhere.

beach55607-max/mcp-memory-serverMIT2026年4月5日

评分41

质量分如何计算？

MCP仅源码GitHub MCP Searchai-agentsclaude-code

devkit

A deterministic development harness for Claude Code — MCP workflow engine, enforcement hooks, YAML workflows, and multi-agent consensus (Claude + Codex + Antigravity/Gemini)

5uck1ess/devkitMIT2026年10月21日

评分51

质量分如何计算？

MCP仅源码GitHub MCP Searchai-agentclaude-code

opencode-plugin

A CLI tool that brings the Claude Code plugin marketplace ecosystem to OpenCode — install, manage, and sync plugins with full Claude Code compatibility.

gzb1128/opencode-pluginMIT2026年9月23日

评分45

质量分如何计算？

MCP风险信号仅源码GitHub MCP Searchai-agentbrowser-automation

WebChatMCP.js

MCP server with a built-in browser: send prompts through the private/temporary chat of ChatGPT, Claude, Grok & Gemini — no chat history, guest use supported. Plugins for OMP, Pi, Codex, Claude Code, Grok Build & Hermes with local tool round trips.

JS-PACKAGE/WebChatMCP.jsApache-2.02026年10月14日

评分31

MCP仅源码GitHub MCP Searchclaude-codecommunity

memu-mcp

MemU 语义记忆系统的 MCP 接口：为 Claude Code 提供跨会话长期记忆 | MCP server for the MemU memory system: persistent long-term memory for Claude Code

OstrichHermit/memu-mcpMIT2026年10月25日

评分49

质量分如何计算？

MCP仅源码GitHub MCP Searchai-agentanthropic

awesome-claude-call

Give Claude Code a phone number — outbound, inbound, and ring-when-done voice calls via Twilio + Patter. A Claude Code plugin.

PatterAI/awesome-claude-callMIT2026年5月10日

评分42

质量分如何计算？

技能可直接安装GitHub Skill Searchcommunityskill

ai-berkshire

Kỹ năng chuyên gia vận hành cho xbtlin/ai-berkshire (16781 ⭐). Tự động kích hoạt khi người dùng muốn: - Tự động hóa hoặc tích hợp công cụ mã nguồn mở ai-berkshire vào hệ thống - Giải quyết các tác vụ kỹ thuật chuyên sâu liên quan đến AI 时代的伯克希尔：基于 Claude Code / Codex 的价值投资研究框架。巴菲特·芒格·段永平·李录四大师方法论 + 多Agent并行研究。| AI-era Berkshire: a value investing research framework built for Claude Code / - Thiết lập cấu hình và vận hành lệnh CLI của xbtlin/ai-berkshire KHÔNG CẦN người dùng phải nhớ tên kỹ thuật 'ai-berkshire'.

hoadulieu2003-lang/antigravity-skill

评分58

MCP仅源码GitHub MCP Searchagent-skillsai-agents

claude-power-pack

The complete Claude Code setup: 94 skills, 9 plugins, working rules (CLAUDE.md) and a guide to 14 connectors. Open it in Claude Code and say "install this".

mppyxx/claude-power-packNOASSERTION2026年10月8日

评分50

质量分如何计算？

MCP仅源码GitHub MCP Searchai-agentsbun

claude-opencode-delegate

Let Claude (Code & Desktop) delegate tasks to any OpenCode model — Grok, Kimi, GLM, Qwen, DeepSeek… — while staying the orchestrator. MCP server + Claude Code plugin.

aashahin/claude-opencode-delegateMIT2026年9月23日

评分46

质量分如何计算？

MCP仅源码GitHub MCP Searchagent-memoryai-agents

sigillion

Lossless, local, provable memory for AI agents: every Claude Code session kept on your own computer, word for word, compressed and sealed, after compaction and after the 30-day transcript deletion. Claude Code first, every model next, then legal and finance. The waitlist, the releases, the website. Steddy Nova SRL.

Sigillion-AI/sigillionNOASSERTION2026年10月25日

评分55

MCP仅源码GitHub MCP Searchagents-mdai

faf-skills

Claude Code Skills — the FAF Way · curated, receipt-proven skills that take any project to 100% AI-readiness · open SKILL.md format (Claude, Codex, Copilot, Gemini, Cursor)

Wolfe-Jam/faf-skillsMIT2026年10月25日

评分51

质量分如何计算？

MCP仅源码GitHub MCP Searchclaudeclaude-code

argos

Audited SQL access for Claude Code and any MCP client: PostgreSQL, MySQL, SQLite and SQL Server, read-only by default, with field redaction and SSH tunnels. In Claude Code, /argos opens a pane with your schemas, every query's full result and a read-only editor.

AraneaDev/argosMIT2026年10月26日

评分51

质量分如何计算？

MCP仅源码GitHub MCP Searchagentic-aiagentic-engineering

claude-colony

Multi-agent Claude Code you can actually control. Spawn specialist teams in tmux, watch them work side-by-side, steer when needed. Built on native Claude Code agents. No black boxes.

MakingJamie/claude-colonyMIT2026年1月5日

评分42

质量分如何计算？

MCP仅源码GitHub MCP Searchaiai-agent

agent-cost-mcp

MCP server for tracking AI agent costs in real time. Per-message spending, budget alerts, visual dashboard. Works with Claude Code, Cursor, Windsurf, Codex, Gemini CLI, OpenClaw. Stop overspending on AI APIs.MCP server that tracks AI agent spending in real time. Visual dashboard, budget alerts, 15+ models. Works with Claude Code, Cursor, Windsurf.

vanthienha199/agent-cost-mcpMIT2026年3月27日

评分38

MCP仅源码GitHub MCP Searchagnes-aiai-image-generation

media-gen-mcp

Claude Code 的图像全家桶 MCP——AI 生图/视频 + 图像识别(OCR/表格/图表/VLM)+ 架构图/数据图/分享卡/二维码/公式/嵌套可下钻架构图,21 工具一个装好,纯免费。 | All-in-one image MCP for Claude Code — AI image/video gen + recognition (OCR/table/chart/VLM) + diagrams/charts/cards/QR/formula + nested drill-down architecture. 21 tools, free.

wangdong233/media-gen-mcp2026年9月25日

评分44

质量分如何计算？

技能可直接安装GitHub Skill Searchcommunityskill

Claude Prompt Drift Check

Detect drift in the Claude system prompt that agent-container sends to Anthropic. Captures two axes — the bare `claude_code` preset (Anthropic baseline) and agent-container's fully custom wire request — at the SDK version pinned in `agent-container/package-lock.json`, then stores versioned snapshots in `<skill>/snapshots/` (gitignored). Because SuperAgent REPLACES the preset with its own system prompt, preset drift never reaches our sessions automatically — every pure-claude change must be triaged for porting into system-prompt.md. Use when bumping `@anthropic-ai/claude-agent-sdk`, debugging unexpected model behavior after an SDK upgrade, previewing what a PR changes in the on-wire prompt, or auditing what Anthropic silently changed in the Claude Code preset.

SkillfulAgents/SuperAgent