monday.com MCP - AI Developer Tools Tool

Overview

monday.com MCP is an open-source framework from the monday.com AI team that connects AI agents to monday.com’s Work OS using the Model Context Protocol (MCP). It includes a plug-and-play monday API MCP server package and an Agent Toolkit for building agent-driven automations, custom workflows, and integrations that can read and modify structured data (boards, items, columns) via GraphQL and higher-level tools. The project supports both a hosted MCP (recommended for production use and OAuth-managed access) and a local/dev mode for customization and offline development. ([github.com](https://github.com/mondaycom/mcp)) The repo exposes a rich set of prebuilt tools (item/board operations, user/team queries, form creation) plus a beta Dynamic API Tools feature that lets agents generate and run arbitrary GraphQL queries against the monday.com API. Typical uses include natural-language task creation, multi-board reporting, automated item updates, and app lifecycle operations (create/promote apps and check deployments) when using the monday apps MCP. The MCP is presented as compatible with popular agent environments (Claude, Cursor, Copilot, ChatGPT integrations may vary), and monday provides hosted endpoints to avoid local setup. ([github.com](https://github.com/mondaycom/mcp))

GitHub Statistics

• Stars: 375
• Forks: 54
• Contributors: 29
• License: MIT
• Primary Language: TypeScript
• Last Updated: 2026-02-23T16:21:30Z

The GitHub repository is actively maintained by monday.com and licensed under MIT. The public repo shows 375 stars and 54 forks and includes ongoing activity (dozens of commits, open issues and PRs) indicating active development and responsiveness. The project provides detailed docs, examples, and an explicit local development workflow (clone → yarn install → yarn build) and npx-based quick-start commands for clients. According to the repository and project pages, the codebase is intended for contribution and customization. ([github.com](https://github.com/mondaycom/mcp))

Installation

Install via npm:

MONDAY_TOKEN=your_monday_api_token npx @mondaydotcomorg/monday-api-mcp@latest

MONDAY_TOKEN=your_monday_api_token npx @mondaydotcomorg/monday-api-mcp@latest --enable-dynamic-api-tools true

git clone https://github.com/mondaycom/mcp.git

cd mcp && yarn install && yarn build

Configure your MCP client (e.g., Cursor, Claude Desktop) to run: npx @mondaydotcomorg/monday-api-mcp@latest with MONDAY_TOKEN in env

gemini extensions install https://github.com/mondaycom/mcp (optional Gemini CLI extension bundle)

Key Features

• Hosted MCP endpoint for OAuth-managed, no-setup access to monday workspaces.
• Agent Toolkit (agent-toolkit) utilities for building OpenAI/MCP agents quickly.
• Prebuilt tools: create_item, change_item_column_values, create_update, move_item_to_group.
• Board and schema operations: create_board, get_board_schema, create_column, delete_column.
• Account tools: list_users_and_teams and workspace-scoped access controls.
• Dynamic API Tools (beta): generate and execute arbitrary GraphQL queries/mutations.
• monday apps MCP: create_app, promote_app, get_deployment_status for app lifecycle control.

Community

Community engagement is active: the official repo and monday developer docs provide examples and setup guides; GitHub shows active commits, issues, and PRs. monday.com support and developer pages mark MCP as Beta and list integrations (Claude, Cursor, Copilot), and community threads report successful Claude integrations and occasional compatibility or stability issues with some agent clients (e.g., ChatGPT tool-calling differences). Contribution is encouraged via GitHub and community forums. For details and reported experiences, see the official repo, support docs, and community posts. ([github.com](https://github.com/mondaycom/mcp))

Key Information

• Category: Developer Tools
• Type: AI Developer Tools Tool