ClickUp MCP: Connect your AI assistant to your ClickUp workspace

by Sujay ChoubeyJul 17, 20269 min read
AI Agents

TL;DR:

  • You can connect Claude Code, Cursor, or ChatGPT Work to your ClickUp workspace using Composio's managed MCP gateway and CLI.

  • Composio handles OAuth authentication and token refresh automatically, so your connection stays live between sessions.

  • Composio's free tier gives you 20,000 tool calls per month with no credit card required.

  • Once connected, your AI assistant can create tasks, update statuses, and query project data directly from a natural language prompt.

Every time you manually copy a task description from your AI chat into ClickUp, you lose the head start AI already gave you. The context is already in the conversation. The workspace is already open. The only thing missing is a direct connection between the two. The Model Context Protocol (MCP) closes that gap.

Think of MCP as a universal connector between your AI assistant and the apps you use. It standardizes how AI clients send requests to external tools, so the connection works the same way regardless of which AI client or which app you're using.

Connecting your AI assistant to ClickUp shouldn't require hours of debugging API tokens. By using MCP and Composio, you can establish a secure, direct link between your workspace and your AI stack in one sitting. This guide walks you through the exact steps.

How to integrate ClickUp MCP with ChatGPT Work

Prerequisites

  • A ChatGPT Work account

  • The ChatGPT Work desktop app

  • A ClickUp account

  • Composio MCP

Step 1: Open the MCP settings

In ChatGPT Work, open Settings and select Plugins. Click the MCP tab inside the Plugins page.

Step 2: Add the server

Click Add server, enter Composio as the name, select Streamable HTTP, and paste https://connect.composio.dev/mcp into the URL field. Click Save to add the server.

Step 3: Authenticate

Click Authenticate, log in through Composio, and select Allow access. You can then connect your ClickUp account when ChatGPT prompts you to authorize the integration.

How to integrate ClickUp MCP with Claude Code

Prerequisites

  • A Claude account with Claude Code installed

  • A ClickUp account

  • A Composio account

  • Access to your terminal

Step 1: Add the Composio MCP to Claude

Open the terminal and generate the Composio MCP URL. Use the generated URL to add Composio as an MCP server in Claude Code.

Step 2: Start Claude Code

Run the following command in your terminal:

claude

Step 3: Open your MCP list

Enter the following command inside Claude Code to view your connected MCP servers:

/mcp

Step 4: Authenticate Composio

Select Composio from the MCP list and click Authenticate. Claude Code will redirect you to the Composio OAuth page.

Step 5: Connect your ClickUp account

Complete the OAuth flow by authorizing Composio. Then ask Claude to connect to your ClickUp account and authenticate using the link it provides.

How to integrate ClickUp MCP with Cursor

Prerequisites

  • A Cursor account

  • A ClickUp account

  • Access to your project files or global Cursor configuration

  • A Composio account

Step 1: Install Composio in Cursor

Use the Install in Cursor button to add Composio with one click. This is the fastest setup option.

Step 2: Add Composio manually

Open .cursor/mcp.json in your project root, or ~/.cursor/mcp.json for a global configuration, and add:

{
  "servers": {
    "composio": {
      "type": "http",
      "url": "https://connect.composio.dev/mcp"
    }
  }
}

Step 3: Authorize Composio

Restart Cursor, open the MCP Tools settings, and click Connect next to Composio. Complete the authorization flow in the browser window that opens.

Key capabilities enabled by ClickUp MCP

Connecting ClickUp to an AI agent lets you manage project work through simple instructions instead of opening tasks, applying filters, updating fields, and copying information by hand.

The AI agent can read live task data, make changes inside ClickUp, and use the results of one action to decide what to do next. This makes the integration useful for everyday task management, project reporting, deadline tracking, team coordination, and multi-step workflows.

Automate task updates with ClickUp MCP

Sample prompt:

”In MyWorkspace / Product / Sprint-12 / In Progress, change the status of ‘API Error Handling’ to Done and add a comment: ‘Completed and tested on staging. Ready for review.’”

Instead of opening the task, changing its status, and adding a comment manually, you can ask the AI agent to handle the full update.

The agent finds the correct task, marks it as complete, adds the comment, and confirms that the changes were made. This is especially useful when you need to update several task fields at once.

For example, you could ask:

”Find the task called ‘API Error Handling,’ mark it as Done, assign it to Jordan for review, move the due date to July 18, and add a handoff comment.”

The AI agent can complete each action in sequence without requiring you to open ClickUp or edit every field separately.

This can help teams:

  • Update tasks faster after meetings or work sessions

  • Reduce repetitive admin work

  • Keep task statuses and comments current

  • Complete several related updates with one instruction

  • Standardize handoffs between team members

Query your workspace with ClickUp MCP

Sample prompt:

”In MyWorkspace / Engineering / Q3-Roadmap, which tasks have a due date before July 15th and are still marked as Open?”

The AI agent can search your ClickUp workspace and return only the tasks that match your request. It can include useful details such as the task owner, due date, priority, current status, and task link.

This means you do not need to build a new ClickUp filter every time you want to answer a specific question.

You could ask questions such as:

”Which high-priority tasks are overdue?”

”What is Priya currently working on?”

”Which tasks are due this week but do not have an assignee?”

”Show me all blocked tasks in the Website Launch project.”

The agent can also use the results in a follow-up action. After finding overdue tasks, for example, you could ask it to draft a team update, change their priority, or add a comment asking each owner for a revised delivery date.

This makes the integration useful for:

  • Preparing sprint and planning meetings

  • Checking deadlines and project risks

  • Reviewing team workloads

  • Finding unassigned or inactive tasks

  • Turning live project data into summaries and reports

Review active ClickUp project data

Sample prompt:

”Give me a status summary of all tasks in MyWorkspace / Client Projects / Acme Corp that were updated in the last 7 days.”

The AI agent can review recent activity across a ClickUp project and turn it into a clear update. Instead of manually checking each task, it can group the information into completed work, active tasks, blockers, overdue items, and upcoming deadlines.

For example, the summary could show:

Completed

  • Landing page copy approved

  • Analytics tracking implemented

  • Product screenshots delivered

In Progress

  • Homepage redesign, assigned to Alex

  • CRM integration, assigned to Priya

Blocked

  • Pricing page waiting for client approval

  • Migration task waiting for API credentials

You can make the request more specific depending on who will read the update:

”Summarize completed work, current priorities, blockers, and anything that needs client input.”

”Create an internal project summary that includes overdue tasks and owner names.”

”Draft a client-facing update that leaves out internal comments.”

The AI agent can turn the same ClickUp data into different formats for a client report, weekly review, Slack update, or leadership summary.

This can reduce the time spent:

  • Reviewing individual task updates

  • Collecting information from multiple lists

  • Formatting weekly status reports

  • Separating internal details from client-facing updates

  • Identifying tasks that need attention

Combine multiple ClickUp actions in one workflow

The biggest benefit of connecting ClickUp to an AI agent is that you are not limited to one action at a time.

A single request can include searching, filtering, updating, commenting, assigning, and summarizing.

For example:

”Find all overdue tasks in MyWorkspace / Marketing / Website Launch. Group them by assignee, change their priority to Urgent, add a comment asking for a revised delivery date, and give me a summary of the changes.”

The AI agent can:

  1. Find the relevant tasks

  2. Identify which ones are overdue

  3. Group them by owner

  4. Update their priority

  5. Add a comment to each task

  6. Summarize what it changed

This turns a process that may normally require several ClickUp views and repeated manual edits into one instruction.

Other useful workflows include:

  • Turning meeting notes into new ClickUp tasks

  • Creating tasks from emails or support requests

  • Reviewing overdue work and notifying task owners

  • Preparing weekly project summaries automatically

  • Reassigning work when a team member is unavailable

  • Creating follow-up tasks from blockers or unresolved comments

  • Checking whether projects are on track before a client meeting

  • Updating multiple related tasks after a launch or release

The integration goes beyond answering questions. It can work with live project data, make approved changes, and carry out complete workflows across your ClickUp workspace.

Start with Composio's free tier at 20,000 tool calls per month and no credit card required. Run pip install composio and composio connected-accounts link clickup to get the connection live.

FAQs

What AI tools work with ClickUp MCP via Composio?

Claude Code, Cursor, and ChatGPT Work all work with Composio's MCP gateway. Any MCP-compatible client that accepts a server URL configuration can connect using the same Composio endpoint.

How does Composio keep my ClickUp data private?

Composio applies zero-day log retention by default and encrypts all data at rest and in transit. Your workspace data passes through the MCP gateway for tool execution but is not stored beyond immediate use.

What does a ClickUp MCP setup cost?

Composio's free tier includes 20,000 tool calls per month at no cost with no credit card required. The $29/month plan covers 50,000 tool calls. ClickUp API access is available on all plans, including Free Forever, with rate limits varying by tier.

What if ClickUp changes its API and breaks the connection?

Because Composio manages the integration layer, API changes on ClickUp's side become Composio's problem to fix, not yours. You run the CLI setup once. Composio absorbs versioning changes, deprecations, and provider-specific quirks on an ongoing basis, which is the core operational advantage over building and owning a custom integration.

Glossary

MCP (Model Context Protocol): A standard that lets AI assistants connect to external tools and data sources through a shared interface. Instead of each AI client building a custom connection per tool, MCP gives you one setup that works across AI clients.

MCP gateway: A hosted server that acts as the connection point between your AI client and external tools. Composio's MCP gateway sits at https://connect.composio.dev/mcp and routes tool calls from your AI assistant to the correct integration.

OAuth: An authorization standard that lets you grant an application access to your account without sharing your password. When you connect ClickUp through Composio, the OAuth flow issues a token that Composio uses to act on your behalf.

Token refresh: OAuth access tokens expire after a set period. Token refresh is the process of automatically replacing an expired token with a new one. Composio handles this in the background so your ClickUp connection stays live between sessions.

CLI (Command Line Interface): A text-based interface for running commands in your terminal. Composio provides a CLI (composio) for setting up connections, listing accounts, and managing integrations without using a browser.

Streamable HTTP: An MCP transport method that sends and receives data over standard HTTP connections. Used when adding Composio as an MCP server in ChatGPT Work.

Share