FastAPI MCP SSE

What is FastAPI MCP SSE?

FastAPI MCP SSE is a Server-Sent Events (SSE) implementation built on the FastAPI framework that integrates the Model Context Protocol (MCP). It demonstrates how to embed MCP SSE endpoints (/sse and /messages/) into a full-featured FastAPI web application while keeping standard web routes like /, /about, and /status separate. This project is designed for developers who want to add MCP capabilities – allowing AI models to access external tools and data – into their FastAPI applications.

How to use FastAPI MCP SSE?

You can run the server directly without cloning the repository using uvx --from git+https://github.com/panz2018/fastapi_mcp_sse.git start. For full installation, create a virtual environment with uv venv, activate it, install dependencies with uv pip install -r pyproject.toml, then start the server via python src/server.py or uv run start. The server runs on http://localhost:8000 and provides MCP SSE endpoints at /sse and /messages/ alongside standard web routes. Debugging is done using the MCP Inspector with the command mcp dev ./src/weather.py.

Key features of FastAPI MCP SSE

• Server-Sent Events (SSE) integration with the Model Context Protocol (MCP)
• FastAPI framework with custom route support
• Unified web application combining MCP and standard web endpoints
• Clean separation of concerns between MCP and web functionality
• Modular architecture for easy extension with new routes

Use cases of FastAPI MCP SSE

• Enabling AI models to query real-time weather data through MCP tools
• Adding MCP SSE endpoints to an existing FastAPI web service
• Prototyping and debugging MCP tool integration with MCP Inspector
• Building hybrid applications that serve both standard web pages and MCP-based AI interactions

FAQ from FastAPI MCP SSE

What is the purpose of this server?

It provides a reference implementation for combining FastAPI web routing with MCP’s SSE transport, allowing developers to create FastAPI apps that expose both regular web endpoints and MCP tool capabilities.

What are the prerequisites to run the server?

Install the UV Package Manager (a Python package installer in Rust). Python and the dependencies listed in pyproject.toml are also required.

How can I test the MCP functionality?

Use the MCP Inspector: run mcp dev ./src/weather.py, open the inspector at http://localhost:5173, set the Transport Type to SSE and URL to http://localhost:8000/sse, then connect to list and run tools like get_alerts or get_forcast.