mcp

This MCP server connects Taskade to AI assistants via the Model Context Protocol, enabling you to control Taskade workspaces, projects, tasks, agents, and more from any MCP client such as Claude Desktop, Cursor, Windsurf, or n8n. The server exposes a suite of 50+ tools organized into categories like Workspaces, Projects, Tasks, AI Agents, Templates, and Media, allowing your AI agent to list, create, update, or delete resources, manage assignments, dates, notes, and field values, and even publish or share agents. You can run the server in default mode (SSE) or in HTTP mode for custom clients, with the server typically accessible at http://localhost:3000 and SSE at http://localhost:3000/sse?access_token=YOUR_API_KEY. To get started, supply your Taskade API key and configure the MCP client with the provided mcpServers/taskade configuration. The tooling supports interacting with workspaces, projects, tasks, agents, and more through straightforward tool calls like projectCreate, taskGet, folderAgentGenerate, and agentKnowledgeProjectCreate.

Prerequisites:

• Node.js (recommended, with npm or corepack).
• A Taskade account and a generated API key (PAT) from Taskade Settings > API).

Installation steps:

1. Ensure Node.js and npm are installed:

   • node -v
   • npm -v

2. Run the MCP server via npx as shown in the Quick Start:

   TASKADE_API_KEY=your-api-key npx -y @taskade/mcp-server
   

3. (Optional) Run in HTTP mode for custom clients:

   TASKADE_API_KEY=your-api-key npx @taskade/mcp-server --http
   

4. If you want to configure client integration, create the appropriate mcp_config entry (see the provided example) and adjust the PORT via environment variable if you need a non-default port.

5. Verify the server is running at http://localhost:3000 (default) and that SSE is available at http://localhost:3000/sse?access_token=your_api_key.

Tips and common issues:

• Ensure TASKADE_API_KEY is kept secret and only provided to your MCP client configuration.
• If the server doesn’t start on port 3000, set the PORT environment variable (e.g., PORT=4000) and connect to that port.
• When using HTTP mode, your client will communicate over HTTP(S); ensure network access and CORS considerations are addressed if applicable.
• The mcp_config example uses npx with the @taskade/mcp-server package; you can also pin a specific version if needed (e.g., @taskade/mcp-server@x.y.z).
• The server integrates with Taskade’s API; ensure your API key has appropriate permissions for the actions you intend to perform (read/write on workspaces, projects, tasks, etc.).