Google-Drive-MCP

jayeshchowdary/Google-Drive-MCP

3.2

If you are the rightful owner of Google-Drive-MCP and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Google Drive MCP Server is a comprehensive Model Context Protocol (MCP) server that provides access to Google Drive functionality through a variety of tools for file management, sharing, comments, and more.

Tools
5
Resources
0
Prompts
0

Google Drive MCP Server

A comprehensive Model Context Protocol (MCP) server that provides access to Google Drive functionality through 42+ tools for file management, sharing, comments, and more. This server can be used locally, deployed to Vercel, integrated with Claude Desktop, or tested with MCP Inspector.

🚀 Live Deployment

Production URL: https://gdrive-mcp.vercel.app/api/mcp

📁 Project Structure

gdrive-mcp/
├── api/
│   └── mcp.py              # Vercel API endpoint
├── server.py               # Main FastMCP server with Google Drive tools
├── vercel.json             # Vercel deployment configuration
├── requirements.txt        # Python dependencies
├── pyproject.toml          # Project configuration
├── local_oauth.py          # OAuth setup utility
├── test_authenticated.py   # Test script for authenticated tools
└── README.md              # This file

🛠️ Available Tools

The server provides 42+ Google Drive tools including:

File Management

  • create_file_or_folder - Create files and folders
  • create_file_from_text - Create files from text content
  • download_file - Download files from Drive
  • edit_file - Edit file content
  • delete_folder_or_file - Delete files/folders
  • copy_file - Copy files
  • find_file - Search for files
  • list_files - List files with pagination
  • move_file - Move files between folders
  • trash_file - Move files to trash
  • untrash_file - Restore files from trash

Folder Operations

  • create_folder - Create new folders
  • find_folder - Find folders by name
  • list_file_labels - List file labels

Sharing & Permissions

  • add_file_sharing_preference - Share files
  • list_permissions - List file permissions
  • delete_permission - Remove permissions
  • check_file_permissions - Check file access
  • update_permission - Update file permissions
  • get_permission - Get specific permission details

Comments & Collaboration

  • create_comment - Add comments to files
  • get_file_comments_and_replies - Get all comments
  • delete_comment - Delete comments
  • delete_reply - Delete comment replies
  • update_comment - Update existing comments
  • update_reply - Update comment replies
  • list_comments - List file comments
  • get_comment - Get specific comment
  • get_reply - Get specific reply
  • list_replies_to_comment - List replies to a comment

Google Workspace Integration

  • edit_google_workspace_file - Edit Docs, Sheets, Slides
  • create_shortcut_to_file - Create file shortcuts

Drive Management

  • get_about - Get account and storage info
  • empty_trash - Empty trash
  • get_changes_start_page_token - Track changes
  • list_changes - List file changes
  • get_file_metadata - Get detailed file information
  • update_file_metadata - Update file properties
  • list_revisions - List file revisions
  • generate_ids - Generate unique IDs

Advanced Operations

  • get_drive - Get shared drive information
  • delete_shared_drive - Delete shared drives
  • list_file_labels - List file labels

And many more! See server.py for the complete list.

🔧 Prerequisites

Required Software

  • Python 3.8+ (3.13+ recommended)
  • Google Cloud Console account (for OAuth credentials)
  • Vercel account (for deployment)
  • Claude Desktop (for Claude integration)
  • Node.js (for MCP Inspector)

Required Accounts

  • Google Account with Drive access
  • Google Cloud Console project with Drive API enabled
  • Vercel account (free tier available)

🌐 Google Cloud Console Setup

Step 1: Create Google Cloud Project

  1. Open your web browser
  2. Navigate to: https://console.cloud.google.com/
  3. Sign in with your Google account
  4. Click "Select a project" dropdown at the top
  5. Click "New Project"
  6. Enter project name: Google Drive MCP Server
  7. Click "Create"
  8. Wait for project creation (may take a few minutes)

Step 2: Enable Google Drive API

  1. In the Google Cloud Console, make sure your new project is selected
  2. Click the hamburger menu (☰) in the top-left corner
  3. Navigate to: APIs & Services → Library
  4. Search for: "Google Drive API"
  5. Click on "Google Drive API"
  6. Click "Enable"
  7. Wait for API to be enabled

Step 3: Create OAuth 2.0 Credentials

  1. In the Google Cloud Console, navigate to: APIs & Services → Credentials
  2. Click "Create Credentials"
  3. Select "OAuth client ID"
  4. If prompted, click "Configure Consent Screen"
  5. Choose "External" user type
  6. Click "Create"
  7. Fill in the required fields:
    • App name: Google Drive MCP Server
    • User support email: Your email
    • Developer contact information: Your email
  8. Click "Save and Continue"
  9. Skip "Scopes" (click "Save and Continue")
  10. Add test users (your email address)
  11. Click "Save and Continue"
  12. Review and click "Back to Dashboard"

Step 4: Create OAuth Client ID

  1. Back in Credentials page, click "Create Credentials"
  2. Select "OAuth client ID"
  3. Application type: Select "Web application"
  4. Name: Google Drive MCP Server
  5. Authorized redirect URIs: Add these URLs:
    • https://gdrive-mcp.vercel.app/api/mcp/callback
    • http://localhost:8080/callback (for local testing)
  6. Click "Create"
  7. Copy the Client ID and Client Secret (you'll need these later)
  8. Click "OK"

📦 Local Development Setup

Step 1: Clone and Setup Project

  1. Open Terminal/Command Prompt
  2. Navigate to your desired directory: 
    cd /path/to/your/projects
  3. Clone the repository (if not already done): 
    git clone <repository-url>
cd gdrive-mcp

Step 2: Install Dependencies

  1. Create a virtual environment: 
    python -m venv venv
  2. Activate the virtual environment:
    • On macOS/Linux: 
      source venv/bin/activate
    • On Windows: 
      venv\Scripts\activate
  3. Install dependencies: 
    pip install -r requirements.txt

Step 3: Set Up Environment Variables

  1. Create a .env file in the project root: 
    touch .env
  2. Open the .env file in a text editor
  3. Add your Google OAuth credentials: 
    GOOGLE_CLIENT_ID=your_client_id_here
GOOGLE_CLIENT_SECRET=your_client_secret_here
GOOGLE_REDIRECT_URI=http://localhost:8080/callback
  4. Save the file

Step 4: Run Local OAuth Setup

  1. In Terminal, make sure you're in the project directory
  2. Run the OAuth setup script: 
    python local_oauth.py
  3. Follow the prompts:
    • Copy the OAuth URL that appears
    • Open the URL in your browser
    • Sign in with your Google account
    • Grant permissions to the application
    • Copy the authorization code from the redirect URL
    • Paste the code back into the terminal
  4. Wait for credentials to be saved to .token.json

Step 5: Test Local Server

  1. Run the test script: 
    python test_authenticated.py
  2. Verify output shows successful authentication and tool execution

🌐 Vercel Deployment

Step 1: Install Vercel CLI

  1. Open Terminal/Command Prompt
  2. Install Vercel CLI globally: 
    npm install -g vercel
  3. Verify installation: 
    vercel --version

Step 2: Deploy to Vercel

  1. In Terminal, navigate to the project directory
  2. Login to Vercel: 
    vercel login
  3. Follow the browser authentication (Vercel will open a browser window)
  4. Deploy the project: 
    vercel --prod
  5. Follow the prompts:
    • Set up and deploy? → Type y and press Enter
    • Which scope? → Select your account
    • Link to existing project? → Type n and press Enter
    • What's your project's name? → Type gdrive-mcp and press Enter
    • In which directory is your code located? → Press Enter (current directory)
    • Want to override the settings? → Type n and press Enter
  6. Wait for deployment to complete
  7. Note the deployment URL (e.g., https://gdrive-mcp.vercel.app)

Step 3: Set Up Vercel Environment Variables

  1. Open your web browser
  2. Navigate to: https://vercel.com/dashboard
  3. Click on your project (gdrive-mcp)
  4. Click "Settings" tab
  5. Click "Environment Variables" in the left sidebar
  6. Add the following variables:
    • Name: GOOGLE_CLIENT_ID
      • Value: Your Google Client ID
      • Environment: Production, Preview, Development
    • Name: GOOGLE_CLIENT_SECRET
      • Value: Your Google Client Secret
      • Environment: Production, Preview, Development
    • Name: GOOGLE_REDIRECT_URI
      • Value: https://gdrive-mcp.vercel.app/api/mcp/callback
      • Environment: Production, Preview, Development
  7. Click "Save" for each variable

Step 4: Complete OAuth Setup for Vercel

  1. In Terminal, run the OAuth setup: 
    python local_oauth.py
  2. Follow the prompts to complete OAuth
  3. Copy the OAuth token from the output
  4. In Vercel dashboard, add another environment variable:
    • Name: GOOGLE_OAUTH_TOKEN
    • Value: The OAuth token from step 3
    • Environment: Production, Preview, Development
  5. Click "Save"
  6. Redeploy the project: 
    vercel --prod

Step 5: Test Vercel Deployment

  1. Test the deployment: 
    python test_authenticated.py
  2. Verify the server is working: 
    curl https://gdrive-mcp.vercel.app/api/mcp

🤖 Claude Desktop Integration

Step 1: Install Claude Desktop

  1. Open your web browser
  2. Navigate to: https://claude.ai/download
  3. Download Claude Desktop for your operating system
  4. Install the application
  5. Launch Claude Desktop

Step 2: Create MCP Proxy Script

  1. In your project directory, create a file called mcp_proxy.py: 
    touch mcp_proxy.py
  2. Open the file in a text editor
  3. Copy and paste this content: 
    #!/usr/bin/env python3
"""
MCP Proxy for Google Drive MCP Server
This script acts as a bridge between Claude Desktop and the Vercel MCP server
"""

import urllib.request
import urllib.parse
import json
import sys
import signal

# Vercel MCP Server URL
SERVER_URL = "https://gdrive-mcp.vercel.app/api/mcp"

def handle_request(request_data):
    """Handle a single MCP request"""
    try:
        # Convert request to JSON
        json_data = json.dumps(request_data).encode('utf-8')
        
        # Create HTTP request
        req = urllib.request.Request(
            SERVER_URL,
            data=json_data,
            headers={'Content-Type': 'application/json'}
        )
        
        # Send request
        with urllib.request.urlopen(req, timeout=30) as response:
            result = response.read().decode('utf-8')
            return json.loads(result)
            
    except Exception as e:
        return {
            "jsonrpc": "2.0",
            "id": request_data.get("id"),
            "error": {
                "code": -32603,
                "message": f"Proxy error: {str(e)}"
            }
        }

def main():
    """Main MCP proxy loop"""
    print("MCP Proxy starting...", file=sys.stderr)
    
    # Handle shutdown gracefully
    def signal_handler(sig, frame):
        print("MCP Proxy shutting down...", file=sys.stderr)
        sys.exit(0)
    
    signal.signal(signal.SIGINT, signal_handler)
    signal.signal(signal.SIGTERM, signal_handler)
    
    try:
        # Read requests from stdin
        for line in sys.stdin:
            line = line.strip()
            if not line:
                continue
                
            try:
                # Parse JSON-RPC request
                request_data = json.loads(line)
                
                # Handle the request
                response = handle_request(request_data)
                
                # Send response to stdout
                print(json.dumps(response))
                sys.stdout.flush()
                
            except json.JSONDecodeError as e:
                error_response = {
                    "jsonrpc": "2.0",
                    "id": None,
                    "error": {
                        "code": -32700,
                        "message": f"Parse error: {str(e)}"
                    }
                }
                print(json.dumps(error_response))
                sys.stdout.flush()
                
    except KeyboardInterrupt:
        print("MCP Proxy interrupted", file=sys.stderr)
    except Exception as e:
        print(f"MCP Proxy error: {e}", file=sys.stderr)
        sys.exit(1)

if __name__ == "__main__":
    main()
  4. Save the file
  5. Make it executable: 
    chmod +x mcp_proxy.py

Step 3: Configure Claude Desktop

  1. Open File Explorer/Finder
  2. Navigate to the Claude configuration directory:
    • macOS: ~/Library/Application Support/Claude/
    • Windows: %APPDATA%\Claude\
    • Linux: ~/.config/claude/
  3. Create the directory if it doesn't exist
  4. Create or edit claude_desktop_config.json
  5. Add this configuration: 
    {
  "mcpServers": {
    "google-drive-mcp": {
      "command": "python3",
      "args": ["/Users/jayeshkumarchowdary/Desktop/gdrive-mcp/mcp_proxy.py"],
      "env": {}
    }
  }
}
    Note: Replace the path with your actual project directory path
  6. Save the file

Step 4: Restart Claude Desktop

  1. Close Claude Desktop completely
  2. Reopen Claude Desktop
  3. Wait for it to fully load

Step 5: Test Claude Desktop Integration

  1. Start a new conversation in Claude Desktop
  2. Type this test message: 
    What Google Drive tools are available through the MCP server?
  3. Claude should respond with a list of available tools
  4. Try additional commands: 
    Get my Google Drive account information

    List the first 5 files in my Google Drive

    Search for files containing "test" in my Google Drive

🔍 MCP Inspector Testing

Step 1: Install MCP Inspector

  1. Open Terminal/Command Prompt
  2. Install MCP Inspector globally: 
    npm install -g @modelcontextprotocol/inspector
  3. Verify installation: 
    mcp-inspector --version

Step 2: Start MCP Inspector

  1. In Terminal, run: 
    mcp-inspector
  2. Wait for the browser to open automatically (usually http://localhost:3000)

Step 3: Connect to Your MCP Server

  1. In the MCP Inspector browser window:

    • Look for the "Add Server" button or "+" icon
    • Click "Add Server"

  2. Configure the connection:

    • Server Type: Select "HTTP" or "URL"
    • Server URL: Enter https://gdrive-mcp.vercel.app/api/mcp
    • Name: Enter Google Drive MCP (optional)
    • Click "Connect" or "Add"

Step 4: Test MCP Methods

  1. Initialize the connection:

    • Look for "Initialize" button or method
    • Click "Initialize"
    • You should see a success response with server capabilities

  2. List available tools:

    • Look for "Tools" section or "List Tools" button
    • Click "List Tools" or "tools/list"
    • You should see all 42+ Google Drive tools listed

  3. Test a tool:

    • Find the "get_about" tool in the list
    • Click on it to expand
    • Click "Call Tool" or "Execute"
    • You should see your Google Drive account information

Step 5: Test Additional Tools

  1. Try the "list_files" tool:

    • Click on "list_files"
    • Set page_size to 5
    • Click "Call Tool"
    • You should see a list of files

  2. Try the "find_file" tool:

    • Click on "find_file"
    • Set q to "test"
    • Click "Call Tool"
    • You should see search results

🧪 HTTP API Testing

Step 1: Test Basic Connectivity

  1. Open Terminal/Command Prompt
  2. Test the server status: 
    curl https://gdrive-mcp.vercel.app/api/mcp
  3. You should see: 
    {
  "message": "Google Drive MCP Server is running",
  "status": "working",
  "endpoint": "/api/mcp"
}

Step 2: Test MCP Protocol

  1. Test initialization:

    curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 1, "method": "initialize", "params": {}}'

  2. Test tools listing:

    curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 2, "method": "tools/list", "params": {}}'

  3. Test tool execution:

    curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": {"name": "get_about", "arguments": {}}}'

Step 3: Test with Python Script

  1. Run the test script: 
    python test_authenticated.py
  2. You should see:
    • ✅ get_about successful!
    • ✅ list_files successful!

🔧 Troubleshooting

Common Issues

MCP Inspector Connection Issues
  • Clear browser cache and restart MCP Inspector
  • Verify the URL is exactly https://gdrive-mcp.vercel.app/api/mcp
  • Check if the server is running with a simple curl test
Claude Desktop Not Connecting
  • Check the file path in claude_desktop_config.json
  • Make sure the proxy script is executable
  • Restart Claude Desktop completely
  • Check Claude Desktop logs for error messages
Authentication Errors
  • Verify OAuth token is set in Vercel environment variables
  • Re-run the OAuth setup if token has expired
  • Check Google Cloud Console redirect URI configuration
Tool Execution Errors
  • Check Vercel logs for server-side errors
  • Verify OAuth scopes match Google Cloud Console configuration
  • Test with MCP Inspector first to isolate issues

Debug Commands

Check Server Status
curl -I https://gdrive-mcp.vercel.app/api/mcp
Test OAuth Token
curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "get_about", "arguments": {}}}'
Check Vercel Logs
vercel logs https://gdrive-mcp.vercel.app

📝 Usage Examples

Basic File Operations

# List files
curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "list_files", "arguments": {"page_size": 10}}}'

# Search for files
curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 2, "method": "tools/call", "params": {"name": "find_file", "arguments": {"q": "test"}}}'

# Get file metadata
curl -X POST https://gdrive-mcp.vercel.app/api/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": {"name": "get_file_metadata", "arguments": {"file_id": "your_file_id"}}}'

Claude Desktop Commands

List all files in my Google Drive
Search for PDF files
Create a new folder called "MCP Test"
Download the file with ID "1Di1u4M-jra0JzyXuNjfLPvqkEjF1WeOW"
Share the file "testfile01" with read access
Get my Google Drive storage information

🚀 Advanced Configuration

Custom OAuth Scopes

To add additional Google APIs, update the SCOPES list in server.py:

SCOPES = [
    "https://www.googleapis.com/auth/drive",
    "https://www.googleapis.com/auth/drive.file",
    "https://www.googleapis.com/auth/drive.metadata.readonly",
    "https://www.googleapis.com/auth/spreadsheets",
    "https://www.googleapis.com/auth/documents",
    # Add more scopes as needed
]

Environment-Specific Configuration

For different environments, create separate .env files:

  • .env.local - Local development
  • .env.staging - Staging environment
  • .env.production - Production environment

📄 License

This project is open source and available under the MIT License.

🤝 Contributing

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Add tests if applicable
  5. Submit a pull request

📞 Support

For issues and questions:

  1. Check the troubleshooting section above
  2. Review the Vercel logs for server-side errors
  3. Test with MCP Inspector to isolate client-side issues
  4. Check Google Cloud Console for OAuth configuration issues

🎯 Quick Start Summary

  1. Set up Google Cloud Console (OAuth credentials)
  2. Deploy to Vercel (production server)
  3. Complete OAuth setup (authentication)
  4. Choose your integration:
    • Claude Desktop: Use the proxy script configuration
    • MCP Inspector: Connect to the HTTP endpoint
    • HTTP API: Use curl or your preferred HTTP client
    • Local development: Run the server locally

Your Google Drive MCP server is now ready to use! 🚀