google-mcp — Google Docs, Sheets & Drive MCP Server

A production-ready MCP server for Google Docs, Google Sheets, and Google Drive with OAuth token authentication. Deploy to Railway or run locally with Claude Desktop.

Features

• Google Drive: List, search, create, move, copy, delete files and folders; manage sharing permissions

• Google Docs: Read, create, append, find-and-replace, and insert text into documents

• Google Sheets: Create spreadsheets, read/write ranges, append rows, manage sheets

• OAuth Authentication: Uses access tokens with automatic refresh on expiration

• Dual Transport: HTTP/SSE for claude.ai and Claude Desktop via stdio

Related MCP server: Google Drive MCP Server

Setup

1. Create a Google Cloud Project

1. Go to Google Cloud Console

2. Create a new project

3. Enable these APIs:

   • Google Drive API

   • Google Docs API

   • Google Sheets API

2. Create OAuth 2.0 Credentials

1. Go to Credentials in the Google Cloud Console

2. Click Create Credentials → OAuth 2.0 Client ID

3. Application type: Web application

4. Authorized redirect URIs:

   • https://your-api.up.railway.app/api/oauth/google/callback (if using bridge-ai-ops for OAuth)

   • Or any OAuth callback URL you're using

5. Copy the Client ID and Client Secret

3. Get Google Tokens

Use the bridge-ai-ops-api OAuth flow to get OAuth tokens:

1. POST /api/oauth/start with appId for Google (set up in bridge-ai-ops app_catalog)

2. Get redirected to Google login

3. Authorize access to Docs, Sheets, Drive

4. Tokens are stored and can be exported via the platform

Or manually exchange an auth code:

curl -X POST https://oauth2.googleapis.com/token \
  -H "Content-Type: application/x-www-form-urlencoded" \
  -d "code=AUTH_CODE&client_id=CLIENT_ID&client_secret=CLIENT_SECRET&redirect_uri=REDIRECT_URI&grant_type=authorization_code"

Environment Variables

Tools

Google Drive (9 tools)

Google Docs (6 tools)

Google Sheets (8 tools)

Local Development

npm install
npm run build

# Test locally
GOOGLE_ACCESS_TOKEN="..." GOOGLE_REFRESH_TOKEN="..." GOOGLE_CLIENT_ID="..." GOOGLE_CLIENT_SECRET="..." npm start

Server will run on http://localhost:8080

• Health check: curl http://localhost:8080/health

• Claude Desktop connection: stdio with TRANSPORT=stdio

Claude Desktop Configuration

Remote (Railway HTTP)

{
  "mcpServers": {
    "google": {
      "url": "https://your-app.up.railway.app/sse"
    }
  }
}

Local (stdio)

{
  "mcpServers": {
    "google": {
      "command": "node",
      "args": ["/path/to/dist/index.js"],
      "env": {
        "TRANSPORT": "stdio",
        "GOOGLE_CLIENT_ID": "your_client_id",
        "GOOGLE_CLIENT_SECRET": "your_client_secret",
        "GOOGLE_ACCESS_TOKEN": "your_access_token",
        "GOOGLE_REFRESH_TOKEN": "your_refresh_token"
      }
    }
  }
}

Railway Deployment

1. Create a GitHub repository and push this code

2. In Railway, create a new service → Deploy from GitHub

3. Select this repository

4. Set environment variables in Railway dashboard:

   • GOOGLE_CLIENT_ID

   • GOOGLE_CLIENT_SECRET

   • GOOGLE_ACCESS_TOKEN

   • GOOGLE_REFRESH_TOKEN

5. Generate a domain: Settings → Networking → Generate Domain

6. Use https://your-domain.up.railway.app/mcp for claude.ai

7. Use https://your-domain.up.railway.app/sse for Claude Desktop

License

MIT