# Workspace MCP

> **Editing Genesis app source code instead?** See [Hosted MCP — Genesis App (Beta)](/apis-and-developer/workspace-mcp/genesis-app-mcp.md) for the remote server that writes to your app's source files.

Workspace MCP connects [Claude Desktop](https://claude.ai), [Cursor](https://cursor.sh), [Claude Code](https://claude.com/claude-code), or any MCP-compatible AI tool to your Taskade **workspace content** — workspaces, projects, tasks, agents, and media. It's a small server you run locally that wraps the [REST API v1](/apis-and-developer/comprehensive-api-guide.md), so it has full task read/write access.

## What is MCP?

The [Model Context Protocol](https://modelcontextprotocol.io/) lets AI assistants interact with external tools and data sources. The `@taskade/mcp-server` package exposes Taskade's API as MCP tools your AI client can call.

## Install & run

The server is published as **`@taskade/mcp-server`** on npm. The simplest setup runs it on demand with `npx` — no global install needed:

```bash
npx -y @taskade/mcp-server
```

It authenticates with a [Personal Access Token](/apis-and-developer/authentication.md) supplied via the `TASKADE_API_KEY` environment variable. Get yours from [Settings > API](https://www.taskade.com/settings/api).

## Configure Claude Desktop

Add to your Claude Desktop configuration file:

**macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json` **Windows**: `%APPDATA%\Claude\claude_desktop_config.json`

```json
{
  "mcpServers": {
    "taskade": {
      "command": "npx",
      "args": ["-y", "@taskade/mcp-server"],
      "env": {
        "TASKADE_API_KEY": "your_api_token_placeholder"
      }
    }
  }
}
```

## Configure Cursor

Cursor uses the same configuration shape (in `~/.cursor/mcp.json` or the project's `.cursor/mcp.json`):

```json
{
  "mcpServers": {
    "taskade": {
      "command": "npx",
      "args": ["-y", "@taskade/mcp-server"],
      "env": {
        "TASKADE_API_KEY": "your_api_token_placeholder"
      }
    }
  }
}
```

## HTTP / SSE mode

For clients that connect over HTTP instead of stdio, run the server in HTTP mode:

```bash
TASKADE_API_KEY=your_api_token_placeholder npx @taskade/mcp-server --http
```

The server listens on `http://localhost:3000` (set `PORT` to change it); connect via SSE at `http://localhost:3000/sse`.

{% hint style="warning" %}
HTTP mode accepts the token as a query parameter (`?access_token=…`). Only use this on a trusted local network or behind TLS — never expose it publicly.
{% endhint %}

## Available Tools

Tool names mirror the [REST API v1](/apis-and-developer/comprehensive-api-guide.md) operations the server wraps. Highlights:

| Area           | Tools                                                                                                                                                                                                                                                                                                                           |
| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Workspaces** | `workspacesGet`, `workspaceFoldersGet`, `workspaceCreateProject`                                                                                                                                                                                                                                                                |
| **Projects**   | `projectGet`, `projectCreate`, `projectCopy`, `projectComplete`, `projectRestore`, `projectFromTemplate`, `projectMembersGet`, `projectFieldsGet`, `projectShareLinkGet`, `projectShareLinkEnable`, `projectBlocksGet`, `projectTasksGet`                                                                                       |
| **Tasks**      | `taskGet`, `taskCreate`, `taskPut`, `taskDelete`, `taskComplete`, `taskUncomplete`, `taskMove`, `taskAssigneesGet`, `taskPutAssignees`, `taskDeleteAssignees`, `taskGetDate`, `taskPutDate`, `taskDeleteDate`, `taskNoteGet`, `taskNotePut`, `taskNoteDelete`, `taskFieldValueGet`, `taskFieldValuePut`, `taskFieldValueDelete` |
| **Agents**     | `folderAgentGenerate`, `folderCreateAgent`, `agentGet`, `agentUpdate`, `deleteAgent`, `agentKnowledgeProjectCreate`, `agentKnowledgeMediaCreate`, `agentPublicAccessEnable`, `agentConvosGet`, `agentConvoGet`                                                                                                                  |
| **Media**      | `mediasGet`, `mediaGet`, `mediaDelete`                                                                                                                                                                                                                                                                                          |
| **Personal**   | `meProjectsGet`                                                                                                                                                                                                                                                                                                                 |

{% hint style="info" %}
This local server wraps **v1**, so it does **not** include a prompt-an-agent tool. To prompt agents programmatically, use [`POST /api/v2/promptAgent`](/apis-and-developer/api-v2-reference.md#prompt-an-agent), or chat with the agent inside Taskade.
{% endhint %}

## Example Usage in Claude

Once configured, you can ask Claude to:

* "List all my Taskade workspaces"
* "Create a task in my Sales Pipeline project to follow up with the client"
* "Show me the tasks in my Sales Pipeline project and mark the first one complete"
* "Add a due date of next Friday to task X"

## Resources

| Resource                                                                                | Description                                   |
| --------------------------------------------------------------------------------------- | --------------------------------------------- |
| [github.com/taskade/mcp](https://github.com/taskade/mcp)                                | MCP server source code and docs               |
| [Workspace MCP — Advanced](/apis-and-developer/workspace-mcp/workspace-mcp-advanced.md) | Multi-client setup, troubleshooting, security |
| [REST API v1 Reference](/apis-and-developer/comprehensive-api-guide.md)                 | The endpoints these tools map to              |


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.taskade.com/apis-and-developer/workspace-mcp.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.