Google Form MCP Server

🚀 Google Forms Integration Project with CamelAIOrg

This project demonstrates how to integrate Google Forms with the CamelAIOrg agent service through the Model Context Protocol (MCP). Users can create forms via natural language requests, and the system interacts internally using the MCP protocol.

🚀 Quick Start

Using Docker

docker-compose up --build

This command starts the MCP server (listening on port 5000) and the agent service (listening on port 5001).

Environment Configuration

Create a .env file in the project root directory:

# Google API Credentials
GOOGLE_CLIENT_ID=your_client_id
GOOGLE_CLIENT_SECRET=your_client_secret
GOOGLE_REFRESH_TOKEN=your_refresh_token

# Server Configuration
FLASK_ENV=development
PORT=5000
DEBUG=True

# CamelAIOrg Agent Configuration
AGENT_ENDPOINT=http://agents:5001/process
AGENT_API_KEY=your_agent_api_key

Steps to obtain the refresh token:

1. Use the OAuth 2.0 Playground tool: https://developers.google.com/oauthplayground/
2. Configure the project credentials.
3. Filter the required permissions:
   • https://www.googleapis.com/auth/forms
   • https://www.googleapis.com/auth/drive
   • https://www.googleapis.com/auth/drive.file

✨ Features

User Interface

• Submit natural language requests (e.g., "Create a customer feedback form with 3 questions").
• View the request flow between components.
• View detailed information and links of the generated form.

Internal Interaction

• The agent service parses natural language and generates MCP requests.
• The MCP server processes form creation requests and communicates with the Google API.
• The request results are returned to the agent service via the MCP protocol.

Supported Scenarios

• Customer feedback collection
• Market research questionnaires
• Event RSVP forms
• Any custom question types

📦 Installation

Using Docker

docker-compose up --build

💻 Usage Examples

Web Interface Access

Access the address: http://localhost:5000 Interface functions:

1. Enter natural language requests.
2. View the interaction flow between components.
3. View information and links of the generated form.
4. View MCP protocol packet exchange records.

API Interface Description

MCP Server Interface (Port 5000)

• GET /api/health: Health check
• GET /api/swagger: View API documentation
• POST /api/create-form: Create a new form
  • Supports JSON format request body, e.g.:

    {
      "form_name": "Customer Feedback Form",
      "questions": [
        { "type": "Text", "prompt": "Your Name" },
        { "type": "Email", "prompt": "Your Email Address" }
      ]
    }
    

Agent Service Interface (Port 5001)

• POST /process: Process natural language requests
  • Supports HTTP POST requests, e.g.:

    curl -X POST http://localhost:5001/process \
         -H "Content-Type: application/json" \
         -d '{"command": "Create a form", "details": {"question_count": 3}}'
    

📚 Documentation

Project Architecture

google-form-mcp-server/
├── server/                 # MCP server implementation
│   ├── app.py              # Main Flask application
│   ├── config.py           # Configuration management
│   ├── forms_api.py        # Google Forms API integration
│   ├── mcp_handler.py      # MCP protocol processing logic
│   ├── requirements.txt    # Python dependencies
│   ├── static/             # Static resource files
│   ├── templates/          # HTML templates
│   └── utils/              # Utility functions
├── agents/                 # CamelAIOrg agent service
│   ├── agent_integration.py # Agent integration implementation
│   ├── agent_server.py     # Agent API server
│   └── requirements.txt    # Python dependencies
├── Dockerfile              # MCP server Docker image configuration
├── docker-compose.yml      # Multi-container deployment configuration
├── .env.example            # Environment variable example file
└── README.md               # Project documentation

🔧 Technical Details

Project Dependencies

• Python 3.8+
• Flask framework
• google-auth library
• requests library

Contribution Guidelines

1. Format the code (black .) before submitting.
2. Include unit tests when submitting a PR.
3. Report any deficiencies in existing features in the Issues.

📄 License

This project is licensed under the MIT License. See the LICENSE file for specific terms.

Acknowledgments

Thanks to the support of the following projects:

• Model Context Protocol (MCP) community
• CamelAIOrg development team
• Google Developer Services