Task

You can run Task MCP to manage your Dida365 tasks through the MCP protocol. This server exposes a focused API surface for querying, creating, updating, completing, and removing tasks, with convenient date filters, project matching, and proper timezone handling to keep your tasks in sync with Asia/Shanghai time. It is designed to be used by AI assistants and other MCP clients to automate your task workflow.

How to use

Configure your MCP client to connect to the Task MCP endpoint at the specified MCP URL. Once connected, you can perform common task operations through the five core tools exposed by the server: get_tasks, create_task, update_task, delete_task, and complete_task. Each tool accepts parameters that help you filter results, define task attributes, and control task state.

How to install

Prerequisites you need before installation: Node.js and a package manager (npm or Yarn) installed on your system.

Step 1: Install dependencies in your project directory.

yarn install
# or
npm install

Step 2: Prepare environment configuration by copying the example file and editing it.

cp .env.example .env

Step 3: Edit the environment file to provide your Dida365 access token and server port.

# In your .env file, set:
DIDA_ACCESS_TOKEN=your_access_token_here
PORT=3010

Step 4: Start the MCP server in development or production mode.

# Development with hot reload
yarn dev

# Production
yarn start

How to connect from an MCP client

Use the following MCP server URL to connect your client: http://localhost:3010/mcp. Ensure your client is configured to communicate with this URL using the MCP protocol and that the environment is ready to supply the access token when required.

Configuration and endpoints

The server exposes an MCP endpoint at /mcp and a health check at /health. It uses a Fetch-based API client to interact with the Dida365 service and handles time zone conversions between UTC and Asia/Shanghai automatically.

Additional notes

If you need to adjust date filtering, you can use modes such as all, today, yesterday, or recent_7_days when listing tasks. For task creation and updates, you can specify start_date and due_date in formats like YYYY-MM-DD or YYYY-MM-DD HH:MM:SS. Time zone handling ensures dates align with your local time while stored as UTC.

Example MCP client configuration

You can configure your MCP client with an MCP server entry that points to the Task MCP endpoint.

{
  "mcpServers": {
    "task_mcp": {
      "url": "http://localhost:3010/mcp",
      "args": []
    }
  }
}

Available tools

get_tasks

Fetch a list of tasks with optional filters such as mode, keyword, priority, project_name, and completion status.

create_task

Create a new task with mandatory title and optional content, priority, project_name, dates, all-day flag, and reminder.

update_task

Update properties of an existing task identified by task_id, including title, content, priority, project_name, dates, and status.

delete_task

Delete a specific task identified by task_id.

complete_task

Mark a specific task identified by task_id as completed.