FastMCP Google OAuth Template

A minimal FastMCP server template with:

• Google OAuth authentication
• Encrypted persistent storage (survives Railway deploys)
• Ready to deploy on Railway

Quick Start

1. Copy the template

cp .env.example .env

Edit .env with your values.

2. Install dependencies

uv sync

3. Run locally

uv run python main.py

OAuth data is persisted to ./oauth_data/ automatically.

Google OAuth Setup

1. Go to Google Cloud Console → APIs & Services → Credentials
2. Create an OAuth 2.0 Client ID (Web application)
3. Add authorized redirect URIs:
   • https://your-service.up.railway.app/auth/callback
   • http://localhost:8000/auth/callback (for local dev)
4. Copy the Client ID and Client Secret into your .env

Railway Deployment

First-time setup

1. Create a Railway service pointing to this repo.

2. Add a Volume to the service:

   • Mount path: /data
   • Size: 1 GB (~$0.25/month)

3. Set environment variables in Railway:

   Variable	Value
   FASTMCP_JWT_SIGNING_KEY	Long random string — never change in production
   GOOGLE_CLIENT_ID	From Google Cloud Console
   GOOGLE_CLIENT_SECRET	From Google Cloud Console
   BASE_URL	https://your-service.up.railway.app
   OAUTH_STORAGE_PATH	/data/oauth

4. Deploy. OAuth credentials will now persist across deploys.

Why the volume?

FastMCP stores OAuth client registrations and tokens on disk. Without a persistent volume, every Railway deploy creates a fresh filesystem — clients (e.g. Claude) would need to re-authenticate after every deploy.

Adding to Claude

In Claude's MCP settings, add:

https://your-service.up.railway.app/mcp

Adding Tools

Edit main.py and add tools below the marked section:

@mcp.tool
def my_tool(param: str) -> str:
    """Description shown to the AI."""
    return f"Result for {param}"

Environment Variables Reference