简介
awesome-mcp-servers 适合作为 MCP 发现入口:先按需求搜索,再对仓库做验证与审计,最后把少量稳定 server 固化成生产可用的集合。
最适合: 在搭建 MCP 工具栈、需要“发现 + 评审”流程的人
适配: 各种 MCP 客户端(Claude Desktop/Claude Code/Cursor 等)
配置时间: 5–20 分钟
关键事实(已验证)
- 通过 ripgrep 搜索清单可快速得到候选短名单。
- 把短名单变成 allowlist:装之前先做许可证与活跃度检查。
- GitHub:675 stars · 383 forks;最近更新 2026-05-13(GitHub API 验证)。
正文
把“目录”变成可运营标准
精选 5–10 个核心 servers,记录许可证/鉴权/传输方式,并把 .mcp.json 纳入版本控制。
README 原文节选(verbatim)
Awesome MCP Servers 
A comprehensive, community-curated collection of [Model Context Protocol (MCP)](https://modelcontextprotocol.io) servers — covering AI assistants, browser automation, databases, cloud platforms, developer tools, and much more.
What is MCP? The Model Context Protocol is an open standard that lets AI models securely connect to external tools, APIs, databases, and filesystems. Think of it as a USB-C port for AI — a universal interface that any model can use to interact with the world.
Coverage
This repo currently includes coverage for 7260 MCP servers as of May 30, 2025. Due to the limited space, we only show the most recent 30 servers in each category.
Contributing
We welcome submissions! To add your MCP server:
- Fork this repository
- Add your server to the relevant category in
README.md, following the existing format:- [Server Name](https://github.com/owner/repo) - Brief one-sentence description. - Open a pull request — we review regularly
Guidelines:
FAQ
能直接照着清单安装吗? 答:建议先短名单,再验证仓库活跃度、许可证与安全姿态后安装。
如何避免工具面膨胀? 答:维护 allowlist,并对新增 server 强制 review。
最快怎么找 server?
答:用 rg 在 README 里搜你需要的集成关键词。