mondaycom/mcp

Enable AI agents to work reliably - giving them secure access to structured data, tools to take action, and the context needed to make smart decisions.

What it solves

This project provides a framework for connecting AI agents to the monday.com work operating system. It eliminates the need for developers to build complex, custom integrations from scratch by providing a standardized way for agents to access structured data, perform actions, and gain the context necessary to manage projects and workflows.

How it works

The project consists of two primary components:

  1. monday API MCP Server: A plug-and-play server implementation of the Model Context Protocol (MCP), allowing AI agents (like Claude, ChatGPT, or Cursor) to interact with the monday.com API.
  2. Agent Toolkit: A set of utilities for building custom AI agents that support both OpenAI and MCP implementations.

Users can deploy this via a hosted service for immediate use or run it locally using Node.js. It provides a set of predefined tools for item, board, account, and form operations, and a beta "Dynamic API Tools" feature that allows agents to generate and execute custom GraphQL queries to access the full API surface.

Who it’s for

  • AI Agent Developers: Those building assistants, automations, or custom integrations for monday.com.
  • Project Managers: Users who want to integrate AI assistants into their project tracking and workflow management.
  • App Developers: Those using the monday.com Apps Framework to build custom AI-powered apps.

Highlights

  • Broad Compatibility: Ready for use with Claude Desktop, Cursor, ChatGPT, MS Copilot, and Mistral AI.
  • Dynamic API Access: Beta tools that allow agents to explore the GraphQL schema and execute any API endpoint.
  • Flexible Deployment: Offers both a hosted MCP service for ease of use and a local installation for customization.
  • Comprehensive Toolset: Includes specific tools for managing boards, items, and users, as well as tools for the Apps Framework.