MCP FastAPI Server

Overview

This project is a scalable FastAPI server for handling Model Control Protocol (MCP) requests. It is designed to route requests to different AI model services (such as code generation and debugging), enforce API key security, and provide rate limiting and logging. The server is modular, extensible, and ready for production or research use.

Related MCP server: Xplainable MCP Server

Features

• FastAPI-based: High-performance, async Python web server

• API Key Security: Protects endpoints with API key authentication

• Rate Limiting: Per-key or per-client rate limiting (Redis or in-memory)

• Code Generation & Debugging: Specialized endpoints for codegen and debugging models

• Extensible Routers: Easily add new model types or endpoints

• Comprehensive Logging: Info and error logs for all requests and errors

• Health Checks: Endpoints for service and model health

• Environment-based Configuration: Uses .env and config.py for settings

Project Structure

MCP_SERVER/
├── __init__.py
├── auth.py              # API key management and authentication
├── codegen_router.py    # Endpoints for code generation
├── config.py            # App and environment configuration
├── degubber_router.py   # Endpoints for code debugging
├── main.py              # FastAPI app setup and router inclusion
├── middleware.py        # Custom middleware (rate limiting, logging)
├── models.py            # Pydantic models for requests/responses
├── requirements.txt     # Python dependencies
├── services.py          # Model routing and core logic
├── start_server.py      # Production server startup script
├── test_client.py       # (Legacy) server startup script
├── api_test_script.py   # Script to test all main endpoints

Setup & Installation

1. Clone the repository

2. Install dependencies:

   pip install -r requirements.txt

3. (Optional) Set environment variables in a .env file or your shell (see config.py for options)

Running the Server

• Development:

  python start_server.py

• The server will start on http://localhost:8000 by default.

• API docs available at http://localhost:8000/docs

API Usage

Authentication

• All main endpoints require an API key in the X-API-Key header.

• Example keys are defined in auth.py (e.g., mcp-key-dev-123).

Main Endpoints

• GET / — Root info

• GET /health — Server health

• POST /mcp — General MCP request (routes to appropriate model)

• GET /api/v1/codegen/capabilities — Codegen model capabilities

• POST /api/v1/codegen/generate — Generate code (requires write permission)

• GET /api/v1/codegen/templates — Code templates

• GET /api/v1/codegen/health — Codegen health

• GET /api/v1/debugger/capabilities — Debugger model capabilities

• POST /api/v1/debugger/analyze — Analyze code for bugs/errors

• GET /api/v1/debugger/best-practices — Coding best practices

• GET /api/v1/debugger/health — Debugger health

Example Request (with httpx)

import httpx
headers = {"X-API-Key": "mcp-key-dev-123"}
payload = {
    "model": "aiden-7b",
    "prompt": "Generate a Python function to add two numbers",
    "context": {"language": "python"}
}
resp = httpx.post("http://localhost:8000/api/v1/codegen/generate", headers=headers, json=payload)
print(resp.json())

Testing

• Use api_test_script.py to test all main endpoints automatically:

  python api_test_script.py

• The script prints status and response for each endpoint.

Extending the Project

• Add new models: Implement new handlers in services.py and register them in ModelRouter.

• Add new endpoints: Create new routers (see codegen_router.py, degubber_router.py).

• Change rate limiting: Update or extend middleware.py.

• Change API key logic: Update auth.py.

Contributing

• Fork the repo and submit pull requests.

• Please include tests and update documentation for new features.

License

MIT License (or specify your own)