monday.com MCP integrations – AI agents connecting to the work management platform

monday.com MCP: All Available Tools and Integrations Overview

15. April 20265 min read

TL;DR: „monday.com runs two MCP servers: Platform MCP (boards, items, docs) and Apps MCP (app deployment). Plus native integrations for Claude, Cursor, ChatGPT, Copilot, Gemini, Mistral, Perplexity, and Figma Make. 30+ tools for AI-driven workflows."

— Till Freitag

30-Second Version

monday.com has massively expanded its MCP ecosystem. Two MCP servers (Platform + Apps), 30+ tools for board operations, item management, dashboards, docs, and sprints – plus native integrations for 9 AI clients including Claude, Cursor, ChatGPT, and Gemini. This is the most comprehensive MCP implementation of any work management platform.

Note: With Sidekick 3.0, monday.com uses the same MCP infrastructure internally – every tool in this list is also available to Sidekick.

What Is monday MCP?

The Model Context Protocol (MCP) is the open standard through which AI assistants securely access external platforms. monday.com operates a hosted MCP server at https://mcp.monday.com/mcp that gives AI tools like Claude, Cursor, and ChatGPT direct access to your workspace.

Configuration is dead simple:

{
  "mcpServers": {
    "monday-mcp": {
      "url": "https://mcp.monday.com/mcp"
    }
  }
}

No local setup. No token management. Just connect and go.

The Two MCP Servers

monday.com provides two distinct MCP servers for different use cases:

MCP	Purpose	Audience
Platform MCP	Interact with monday.com data – boards, items, columns, users	Customers, developers, agent builders
Apps MCP	Deploy and manage monday.com apps	App developers

Platform MCP: Read and Write Data

The Platform MCP is the open-source server (GitHub: mondaycom/mcp), maintained by the monday.com AI team. It covers all essential platform operations.

Apps MCP: Build and Deploy Apps

The Apps MCP is designed for developers building monday.com apps. It manages app versions, features, deployments, and environment variables – completely separate from the Platform MCP.

{
  "mcpServers": {
    "monday-apps-mcp": {
      "command": "npx",
      "args": [
        "@mondaydotcomorg/monday-api-mcp",
        "-t", "your_monday_api_token",
        "--mode", "apps"
      ]
    }
  }
}

All Available Platform Tools

Here's the complete list of tools currently provided by the Platform MCP (as of April 2026):

Board Operations

Tool	Function
`get_board_items_page`	Get items from a board with filtering, sorting, and pagination
`get_board_info`	Get comprehensive board metadata including columns, groups, and settings
`create_board`	Create a new board
`board_insights`	Calculate aggregated insights by filtering, grouping, and summarizing
`get_board_activity`	Get board activity logs for a specified time range

Item Management

Tool	Function
`create_item`	Create a new item or sub-item with column values
`change_item_column_values`	Modify column values of an existing item
`create_update`	Add an update (comment/post) to an item
`get_updates`	Get updates including author info, timestamps, replies, and file attachment links

Columns and Groups

Tool	Function
`create_column`	Add a new column to a board
`create_group`	Create a new group in a board
`get_column_type_info`	Get JSON schema and metadata for a column type

Users and Teams

Tool	Function
`list_users_and_teams`	Fetch users and/or teams data
`get_user_context`	Fetch current user info and their relevant boards, folders, workspaces, and dashboards
`create_notification`	Send a notification via the monday bell icon

Workspaces and Folders

Tool	Function
`list_workspaces`	List all workspaces available to the user
`workspace_info`	Get boards, docs, and folders in a workspace
`create_workspace`	Create a new workspace
`update_workspace`	Update workspace name, description, or settings
`create_folder`	Create a new folder in a workspace
`update_folder`	Update folder properties
`move_object`	Move folders, boards, or overviews

Documentation

Tool	Function
`create_doc`	Create documents with markdown in workspaces or attached to items
`read_docs`	Retrieve document content as markdown
`add_content_to_doc`	Add markdown content to an existing document

Forms

Tool	Function
`create_form`	Create a new form with an associated board
`update_form`	Update form settings, appearance, and questions
`get_form`	Get form details by token
`form_questions_editor`	Create, update, or delete form questions

Dashboards and Widgets

Tool	Function
`create_dashboard`	Create a dashboard aggregating data from boards
`create_widget`	Add widgets to dashboards
`all_widgets_schema`	Get JSON schema definitions for all widget types

Search

Tool	Function
`search`	Search for boards, documents, forms, and folders

monday dev Sprints

Tool	Function
`get_monday_dev_sprints_boards`	Discover sprints boards and their task boards
`get_sprints_metadata`	Get sprint details including timeline and status
`get_sprint_summary`	Get AI-generated sprint summary with insights

Advanced API Access

Tool	Function
`all_monday_api`	Execute any GraphQL query or mutation
`get_graphql_schema`	Fetch the monday.com GraphQL schema
`get_type_details`	Get detailed information about a GraphQL type

UI Components

These tools render interactive visualizations in supported AI clients:

Tool	Function
`show-table`	Display an interactive board table
`show-chart`	Render pie or bar charts
`show-battery`	Show progress/status battery indicators
`show-assign`	Display smart assignment suggestions

Native AI Integrations

Beyond the open MCP protocol, monday.com provides pre-configured integrations for nine AI clients:

AI Client	Status
Claude (Anthropic)	✅ Natively supported
Cursor	✅ Natively supported
ChatGPT (OpenAI)	✅ Natively supported
Microsoft Copilot	✅ Natively supported
Microsoft Copilot Studio	✅ Natively supported
Perplexity	✅ Natively supported
Mistral (le Chat)	✅ Natively supported
Gemini CLI	✅ Natively supported
Figma Make	✅ Natively supported

This is remarkable: monday.com supports every relevant AI frontend – from developer tools (Cursor) to consumer assistants (ChatGPT, Gemini) to design tools (Figma Make).

Using Both MCPs Together

For app developers, the greatest value comes from combining both MCP servers:

Read board schemas and item data → Platform MCP
Scaffold, deploy, and promote app versions → Apps MCP
Set environment variables and check deployment logs → Apps MCP
Test app behavior by querying live data → Platform MCP

An AI agent can automate a large portion of the app-building workflow end to end.

MCP and Sidekick 3.0: Shared Infrastructure

The tools in this list also form the foundation for Sidekick 3.0, monday.com's built-in AI assistant. While MCP opens access for external AI agents, Sidekick uses the same infrastructure internally. The result: every new MCP tool automatically extends Sidekick – and vice versa.

This combination makes monday.com the only work management platform with a consistent AI strategy: internal (Sidekick) and external (MCP).

What This Means for Teams

For monday.com Users

The MCP integration transforms monday.com into an AI-driven workspace. Instead of clicking through menus, teams can use natural language to:

Create and manage items and sub-items
Build dashboards with widgets
Move boards between workspaces
Generate sprint summaries
Trigger notifications

For Developers and Agent Builders

The all_monday_api fallback is the game-changer: any GraphQL query can be executed via MCP. This makes monday.com a fully programmable platform for AI agents – without needing to build custom API integrations.

For Our Team

We already use the monday MCP in our own workflows – combined with Claude and Cursor. The integration with Sidekick 3.0 makes the platform even more powerful.

Limitations

Permissions: All tools respect monday.com's security model. Access only to data the user can see.
UI components: show-table, show-chart etc. only work in specific AI clients
Plan dependency: Some features require higher tiers
API version: Controllable via headers – important for breaking changes

Our Take

With two MCP servers, 30+ tools, and 9 native integrations, monday.com has built the most comprehensive MCP implementation in the work management space. This isn't an experiment – it's infrastructure.

The combination of a hosted server (no local setup) and the all_monday_api fallback (any GraphQL operation possible) makes this implementation exemplary. Other platforms offer MCP support – monday.com offers an MCP ecosystem.