MCP ServerHTTP

Asana MCP Server

Asana’s remote MCP lives at mcp.asana.com/mcp and brings tasks, projects, and portfolios into MCP tools. Authenticate with a personal access token from your Asana developer settings, and models can list overdue work, create tasks from bullet points, and summarise what a team completed last week — with the same workspace permissions as your user.

Hosted URL

https://mcp.asana.com/mcp

Suggested model

Claude Sonnet 4.5

MCP Playground runs 30+ models on the same workflow: switch anytime, or use Compare mode to run several in parallel and balance quality vs. cost.

Auth

Asana personal access token: Profile → My Settings → Apps → Developer → personal access token with workspace access you need.

What the Asana MCP server does

How models use it and what it is built for.

The Asana MCP is intended for people who already run projects in Asana: PMs, leads, and ICs. Instead of hand-building filters in the UI, you ask in natural language for your overdue items, a cross-project workload rollup, or a new task under an epic. The server maps those intents onto Asana’s API, so the security boundary is your token and the workspaces you have access to. Combine it with Slack MCP for “turn this thread into Asana tasks” style workflows in Agent Studio multi-server mode.

Tools the Asana MCP server exposes

Typical tools an AI model can call. Exact names vary by version.

  • List and search tasks by project, assignee, due date, and section
  • Create, update, and complete tasks, subtasks, and comments
  • Read project and portfolio structure, including custom fields when exposed

Example prompts to try

Copy any of these into MCP Agent Studio after connecting.

  • List overdue tasks in the “Q2 Launch” project grouped by assignee.

  • Create three tasks under “Design polish” with the titles and due dates I just pasted.

  • Summarise what the Eng team completed in the last 7 days across my workspaces.

Models on MCP Playground

This is not a single-model product: you get the same MCP connection with 30+ models (Claude, GPT, Gemini, DeepSeek, open-weight, and more), you can switch mid-conversation, and you can open Compare mode to run the same prompt against multiple models at once. The card above is a suggested starting point for this server — not the only choice.

Default pick for Asana

Claude Sonnet 4.5

Claude Sonnet 4.5 is strong for multi-section task plans and rewrites. For simple lookups, Haiku 4.5 is cheaper; compare in Agent Studio.

Try the Asana MCP server in your browser

Open MCP Agent Studio with the connection pre-filled. Add your token, pick any of 30+ models, and start chatting — no install required.

Try Asana in Agent Studio

Asana MCP server — FAQ

Common questions about connecting, scoping and using it safely.

What is the Asana MCP server?

It is a hosted remote MCP Asana provides so that AI clients can work with your tasks, projects, and (where your token allows) custom fields, using the same data model as the Asana product API.

How do I create a token for the Asana MCP server?

In Asana, open your profile, go to My Settings, then Apps, then the developer or personal access token area, and create a token. Paste it into your MCP client as the access token for mcp.asana.com/mcp.

Can the model delete or reassign my entire backlog?

It can only do what your user can do, and the model should not bulk-change work without you confirming. Start in a test project, use a separate Asana user if needed, and review every proposed create or update before approving the tool call.

Is Asana MCP only for the browser?

No — any MCP client that can reach the URL with your token works. MCP Playground’s Agent Studio is a convenient in-browser way to test with 30+ models and optional multi-server setups.

What model is best for Asana planning?

Claude Sonnet 4.5 is a good default for triage and long task lists. You can A/B a cheaper model in Compare mode in Agent Studio on a typical weekly planning prompt.

Other MCP servers

More on MCP Playground