Ref
Ref
README
Ref MCP
A ModelContextProtocol server that gives your AI coding tool or agent access to documentation for APIs, services, libraries etc. It's your one-stop-shop to keep your agent up-to-date on documentation in a fast and token-efficient way.
For more see info ref.tools
Setup
There are two options for setting up Ref as an MCP server, either via the streamable-http server (experimental) or local stdio server.
This repo contains the legacy stdio server.
stdio
"Ref": {
"command": "npx",
"args": ["ref-tools-mcp"],
"env": {
"REF_API_KEY": <sign up to get an api key>
}
}
Streamable HTTP (experimental)
"Ref": {
"command": "npx",
"args": [
"-y",
"mcp-remote@0.1.0-0",
"https://api.ref.tools/mcp"
"--header",
"x-ref-api-key:<sign up to get an api key>"
]
}
}
As of April 2025, MCP supports streamable HTTP servers. Ref implements this but not all clients support it yet so the most reliable approach is to use mcp-remote as a local proxy. If you know your client supports streamable HTTP servers, feel free to use https://api.ref.tools/mcp directly.
Note for former alpha users: REF_ALPHA config is still supported. You'll be notified if you need to update.
Tools
Ref MCP server provides all the documentation related tools for your agent needs.
ref_search_documentation
A powerful search tool to check technical documentation. Use this tool whenever you need information about any technical platform, framework, API, service, database, or library. It searches through relevant documentation and finds exactly what you need, down to the specific section of the page.
ref_read_url
A tool to read the full content of a web page. This allows your agent to follow links in documentation and web searches.
ref_search_web (optional)
A fallback web search tool to cover cases when ref_search_documentation doesn't find what you need. It will find links to relevant pages on the web and the ref_read_url tool can be used to read the relevant ones.
We include this tool so that Ref can cover all your search needs in one MCP server but if you prefer another search provider you can disable ref_search_web by setting the DISABLE_SEARCH_WEB environment variable to true or by setting the disable_search_web=false url param in the streamable-http server.
Development
npm install
npm run dev
Running with Inspector
For development and debugging purposes, you can use the MCP Inspector tool. The Inspector provides a visual interface for testing and monitoring MCP server interactions.
Visit the Inspector documentation for detailed setup instructions.
To test locally with Inspector:
npm run inspect
Or run both the watcher and inspector:
npm run dev
Local Development
- Clone the repository
- Install dependencies:
npm install
- Build the project:
npm run build
- For development with auto-rebuilding:
npm run watch
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
