This MCP server is a Python-based implementation of the Model Context Protocol using FastAPI and a streamable HTTP interface. It exposes MCP-compatible tools via a FastAPI application mounted under endpoints (e.g., /echo, /math in the examples) so clients can invoke context-aware operations through HTTP streams. The server is designed to be stateless for HTTP transport, enabling simple scaling and easy integration with MCP clients and OAuth-based authorization if you enable the private server integration. You can run multiple MCP instances in the same FastAPI app to compose richer services and expose distinct tool collections.

Prerequisites:

• Python 3.8+ installed on your system
• npm/yarn (optional, if you plan to use any JavaScript tooling)
• Git (to clone the repository)

Step-by-step:

1. Clone the repository git clone https://github.com/rb58853/simple-mcp-server.git cd simple-mcp-server

2. Create and activate a virtual environment (recommended) python -m venv venv source venv/bin/activate # On Windows use: venv\Scripts\activate

3. Install dependencies pip install -r requirements.txt

4. Run the server (using uvicorn as shown in the configuration)

   This starts the FastAPI app and serves MCP endpoints

   python -m uvicorn src.app:app --host 0.0.0.0 --port 8000

5. Optional: run the OAuth server in parallel (if you enable OAuth integration) python3 src/services/fast_mcp/private_server/oauth_server.py

6. Verify startup Open http://127.0.0.1:8000 in your browser to see the help page or endpoint listing.

Docker (optional):

• Build and run with Docker Compose if provided in the repository: docker compose -f docker-compose.yml up -d --build

Tips and common considerations:

• If you enable OAuth integration, set up the .env file with SUPERUSERNAME and SUPERUSERPASSWORD for initial credentials.
• You can run multiple MCP tools by mounting additional FastMCP servers in the same FastAPI application, enabling richer tool sets and independent session management.
• Ensure your environment variables for OAuth (if used) are configured and that the OAuth server is reachable from the MCP server.
• For production deployments, prefer a reverse proxy (e.g., Nginx) in front of uvicorn and consider enabling TLS termination.
• When debugging, check server logs for MCP tool registration messages and any authorization errors from the token verifier.