MCP.so
登录

🌈 Taskade MCP

@taskade

关于 🌈 Taskade MCP

🤖 Taskade MCP · Official MCP server and OpenAPI to MCP codegen. Build AI agent tools from any OpenAPI API and connect to Claude, Cursor, and more.

基本信息

分类

开发工具

许可证

MIT

运行时

node

传输方式

stdio

发布者

taskade

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "mcp-taskade": {
      "command": "npx",
      "args": [
        "@taskade/mcp-server",
        "--http"
      ],
      "env": {
        "TASKADE_API_KEY": "your-api-key"
      }
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Taskade MCP?

Taskade MCP is a Model Context Protocol server that connects your Taskade workspace to AI assistants like Claude, Cursor, Windsurf, and VS Code. It provides 62 tools for reading and writing workspaces, projects, tasks, AI agents, templates, media, webhooks, and more—all controllable via natural language from your AI client.

How to use Taskade MCP?

Get a Personal Access Token from Taskade Settings > API, then run the server using npx -y @taskade/mcp-server with the TASKADE_API_KEY environment variable. Configure your MCP client (Claude Desktop, Cursor, Windsurf, Cline, VS Code) with the provided JSON snippet. Alternatively, start the server in HTTP/SSE mode with --http for remote or custom clients at http://localhost:3000.

Key features of Taskade MCP

  • 62 tools across workspaces, projects, tasks, AI agents, templates, media, and more
  • Create, update, delete, and manage tasks with due dates, assignees, notes, and custom fields
  • Generate and configure autonomous AI agents, attach knowledge, and publish them publicly
  • List and manage projects, folders, members, and custom fields
  • Subscribe to real-time webhooks and hold live conversations with AI agents (beta)
  • No HTTP client code required—describe what you need in natural language

Use cases of Taskade MCP

  • Ask “Show me all overdue tasks” and get a formatted report from your workspace
  • “Create a support agent trained on our docs project” – generates agent, attaches knowledge, publishes it
  • “Set up next week’s sprint from our template” – copies template, populates tasks, assigns team
  • “Summarize yesterday’s completed work” – reads task history and generates a standup summary
  • Build multi-step automations across 100+ services from your IDE

FAQ from Taskade MCP

What is the difference between this Workspace MCP and the Genesis App MCP?

This package (@taskade/mcp-server) is the Workspace MCP – it runs locally via stdio with a personal token and lets your AI client read/write your projects, tasks, and agents. The Genesis App MCP is a separate hosted surface at https://www.taskade.com/mcp that uses OAuth 2.0 and focuses on editing published app code (requires Business+ plan).

Do I need a Taskade API key? How do I get one?

Yes, you need a Personal Access Token. Go to Taskade Settings > API and create one. The server requires the TASKADE_API_KEY environment variable.

Does Taskade MCP work with remote or custom clients?

Yes. Run with TASKADE_API_KEY=your-key npx @taskade/mcp-server --http to start an HTTP server on port 3000 (configurable with PORT). Connect via SSE at http://localhost:3000/sse?access_token=your-api-key.

Are all 62 tools available on free plans?

Some endpoints, such as workspace listing, are gated for free accounts. Paid plans unlock full access (e.g., automations, project creation from templates, some agent features).

How does Taskade MCP differ from using the Taskade REST API directly?

With the MCP server you run a single npx command instead of writing HTTP client code. You interact in natural language, the AI chains multiple tools intelligently, and error handling is automatic.

评论

开发工具 分类下的更多 MCP 服务器