Model Context Protocol for Raycast
@evanzhoudev
基本信息
关于 Model Context Protocol for Raycast
概览
What is Model Context Protocol for Raycast?
Model Context Protocol for Raycast is a Raycast extension that lets you interact with Anthropic’s Model Context Protocol (MCP) directly within Raycast AI. It enables you to install MCP servers and use the tools they expose in Quick AI or AI Chat.
How to use Model Context Protocol for Raycast?
Add MCP servers by pasting their JSON configuration into the extension’s UI (press Cmd + N in the Manage MCP Servers command) or by editing the mcp-config.json file in the extension’s support directory. You can also write custom local servers in JavaScript by placing an index.js and package.json in a subfolder of the Servers Folder. Once servers are installed, use their tools in Raycast AI by invoking it with the @mcp prefix.
Key features of Model Context Protocol for Raycast
- Install and manage MCP servers through a dedicated UI
- Edit server configurations directly in the config file
- Create and run custom local MCP servers (JavaScript only)
- Automatic
npm installfor local server dependencies - Use MCP tools in Quick AI and AI Chat with
@mcp
Use cases of Model Context Protocol for Raycast
- Give Raycast AI access to your local filesystem via the filesystem server
- Use the sequential thinking server to guide AI reasoning tasks
- Integrate any custom MCP server you write into Raycast AI’s toolset
FAQ from Model Context Protocol for Raycast
How do I add an MCP server?
You can add a server by pasting its JSON configuration into the UI under Manage MCP Servers (press Cmd+N) or by directly editing the mcp-config.json file in the extension’s support directory.
Where is the configuration file located?
Open the Manage MCP Servers command and use the Show Config File in Finder action to locate mcp-config.json. It follows a similar pattern to Claude’s desktop MCP config file.
Can I write my own MCP server?
Yes. Create a subfolder in the Servers Folder, place an index.js file with the correct MCP server code and a package.json. The extension will automatically detect it and run npm install if needed. Only JavaScript is supported.
Do I need to configure API keys for servers?
Some servers require API keys or other parameters. When pasting the server JSON, modify any fields (like allowed directories or API keys) before submitting the form.
What happens if my local server has missing dependencies?
The extension will automatically run npm install in the server’s folder the next time it starts, so you don’t need to manually install packages.
桌面聊天 分类下的更多 MCP 客户端
Dive AI Agent 🤿 🤖
openagentplatformDive is an open-source MCP Host Desktop Application that seamlessly integrates with any LLMs supporting function calling capabilities. ✨
HyperChat
bigsweetpotatostudioHyperChat is a Chat client that strives for openness, utilizing APIs from various LLMs to achieve the best Chat experience, as well as implementing productivity tools through the MCP protocol.
Rijksmuseum MCP Client for Rijksmuseum MCP Server
r-huijtsAn Electron-based desktop application that integrates with the Rijksmuseum MCP server using the Model Context Protocol (MCP). This client enables AI-powered interactions with the Rijksmuseum's art collection through a chat interface.
Cherry Studio
cherryhq🍒 Cherry Studio is a desktop client that supports for multiple LLM providers.
UI-TARS Desktop
bytedanceA GUI Agent application based on UI-TARS(Vision-Language Model) that allows you to control your computer using natural language.
评论