saadhaniftaj/MCP-server-for-Monday.com-DIFY-Automation

🚀 Monday.com MCP Server for Dify.ai Integration

A robust Model Context Protocol (MCP) server that integrates Monday.com with AI agents, featuring both MCP protocol and direct HTTP endpoints.

📋 Overview

This project provides a complete solution for integrating Monday.com with AI platforms like Dify.ai through the Model Context Protocol (MCP). The server includes both MCP endpoints for AI agent integration and direct HTTP endpoints for standalone applications.

🌐 Live Demo

Server URL: https://mcp-server-for-mondaycom-dify-automation-production.up.railway.app/

Interactive Demo: Open demo.html in your browser to test all endpoints

🔧 Features

✅ MCP Protocol Endpoints

• Initialize: POST / with {"jsonrpc": "2.0", "method": "initialize"}
• Tools List: POST / with {"jsonrpc": "2.0", "method": "tools/list"}
• Tool Execution: POST / with {"jsonrpc": "2.0", "method": "tools/call"}
• Notifications: HTTP 204 response (prevents cleanup errors)

✅ Direct HTTP Endpoints

• Update Email: POST /api/monday/update-email
• Search Items: POST /api/monday/search-items
• Update Column: POST /api/monday/update-column
• Get Board Schema: GET /api/monday/board-schema/:boardId

✅ Available Tools

1. monday_email_updater - Update email addresses in Monday.com items
2. get_board_items_by_name - Search for items by name
3. change_item_column_values - Update column values for items
4. get_board_schema - Retrieve board structure and columns

🚨 Dify.ai Platform Issues

Current Status: Dify.ai is experiencing authentication and API issues:

• 401 Unauthorized errors on internal APIs
• 400 Bad Request errors on MCP updates
• Platform authentication problems

Our Solution:

• ✅ Server works perfectly independently
• ✅ Direct HTTP endpoints bypass Dify.ai issues
• ✅ MCP endpoints ready when Dify.ai is fixed
• ✅ Standalone demo available

🚀 Quick Start

1. Test the Server

# Health check
curl https://mcp-server-for-mondaycom-dify-automation-production.up.railway.app/health

# MCP initialize
curl -X POST https://mcp-server-for-mondaycom-dify-automation-production.up.railway.app/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "id": 1, "method": "initialize"}'

# Direct HTTP endpoint
curl -X POST https://mcp-server-for-mondaycom-dify-automation-production.up.railway.app/api/monday/update-email \
  -H "Content-Type: application/json" \
  -d '{"itemName": "Test Task", "email": "test@example.com"}'

2. Interactive Demo

Open demo.html in your browser to test all endpoints with a beautiful UI.

3. Dify.ai Integration (When Platform is Fixed)

Use the MCP endpoint: https://mcp-server-for-mondaycom-dify-automation-production.up.railway.app/

📁 Project Structure

├── dify-workaround-server.js    # Main server with dual endpoints
├── demo.html                    # Interactive demo interface
├── package.json                 # Dependencies and scripts
├── railway.json                 # Railway deployment config
└── README.md                    # This file

🔧 Technical Details

MCP Protocol Compliance

• JSON-RPC 2.0 specification
• Protocol version: 2025-03-26
• HTTP 204 for notifications (prevents cleanup errors)
• Enhanced CORS for cross-origin requests

Server Features

• Express.js with CORS support
• Railway deployment
• Health check endpoint
• Error handling and logging
• Mock Monday.com API responses

Security

• Input validation
• Error sanitization
• CORS configuration for Dify.ai domains
• No authentication required (for demo purposes)

🎯 Use Cases

1. AI Agent Integration

• Dify.ai MCP integration (when platform is fixed)
• Claude/Anthropic MCP integration
• Custom AI agent development

2. Direct Application Integration

• Web applications
• Mobile apps
• Backend services
• Automation scripts

3. Demo and Testing

• Interactive demo interface
• API testing
• Development and debugging

🛠️ Development

Local Development

# Install dependencies
npm install

# Start server
npm start

# Test endpoints
curl http://localhost:3000/health

Deployment

The server is automatically deployed to Railway on push to main branch.

📊 API Reference

MCP Endpoints

Initialize

POST /
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "initialize"
}

Tools List

POST /
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/list"
}

Tool Call

POST /
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "monday_email_updater",
    "arguments": {
      "itemName": "Task Name",
      "email": "user@example.com"
    }
  }
}

Direct HTTP Endpoints

Update Email

POST /api/monday/update-email
{
  "itemName": "Task Name",
  "email": "user@example.com"
}

Search Items

POST /api/monday/search-items
{
  "boardId": 123,
  "term": "search term"
}

Update Column

POST /api/monday/update-column
{
  "boardId": 123,
  "itemId": 456,
  "columnValues": "{\"status\": \"done\"}"
}

Get Board Schema

GET /api/monday/board-schema/123

🔍 Troubleshooting

Dify.ai Issues

• 401/400 errors: These are Dify.ai platform issues, not our server
• Solution: Use direct HTTP endpoints or wait for Dify.ai to fix their platform

Server Issues

• Connection timeout: Check Railway deployment status
• CORS errors: Verify domain is in allowed origins
• JSON parsing errors: Ensure proper Content-Type headers

📞 Support

For issues related to:

• Our MCP server: Check this README and demo
• Dify.ai platform: Contact Dify.ai support
• Monday.com API: Refer to Monday.com documentation

🎉 Success Metrics

• ✅ Server deployed and running on Railway
• ✅ All MCP endpoints working correctly
• ✅ Direct HTTP endpoints functional
• ✅ Interactive demo available
• ✅ Comprehensive documentation
• ✅ Bypass solution for Dify.ai issues

📝 License

MIT License - Feel free to use and modify for your projects.