MCP Server for Windsurf

What is MCP Server for Windsurf?

MCP Server for Windsurf is a Master Control Program server that extends the Windsurf agentic IDE with additional tools not built into the IDE. It uses FastAPI and WebSocket to provide real-time communication and a RESTful API for tool execution.

How to use MCP Server for Windsurf?

Start the server by running python main.py, which launches it on http://localhost:8089. Connect Windsurf to the server by configuring a WebSocket connection to ws://localhost:8089/ws. API documentation is available at http://localhost:8089/docs.

Key features of MCP Server for Windsurf

• FastAPI backend for high-performance API serving.
• WebSocket support for real-time bidirectional communication.
• Extensible tool system using a @register_tool decorator.
• RESTful API endpoints for listing and executing tools.
• Seamless integration with Windsurf IDE.
• Includes built-in tools: File Search, Code Analysis, and Web Request.

Use cases of MCP Server for Windsurf

• Extend Windsurf with custom tools for development workflows.
• Search files with pattern matching directly from the IDE.
• Analyze code files for syntax, complexity, and dependencies.
• Make HTTP requests to external APIs without leaving Windsurf.
• Add new tools by implementing handlers in tools.py.

FAQ from MCP Server for Windsurf

What does MCP Server for Windsurf do?

It provides additional tools for the Windsurf agentic IDE that are not built into the IDE, using a FastAPI server with WebSocket support.

How do I install and run it?

Clone the repository, create a Python virtual environment, install dependencies from requirements.txt, and run python main.py to start the server.

How do I connect Windsurf to the MCP server?

Configure Windsurf to use a WebSocket connection to ws://localhost:8089/ws. The server runs on localhost port 8089.

How can I add custom tools?

Open tools.py and use the @register_tool decorator with a name, description, and parameter schema, then implement the handler function.

What are the runtime requirements?

Python 3 and the dependencies listed in requirements.txt. A virtual environment is recommended for installation.