mcp-task-manager

MCP Task Manager for Cursor

A powerful Model-Context-Protocol (MCP) server for AI task management that integrates with Cursor IDE. This system helps AI assistants structure coding projects into manageable tasks.

Features

• 📋 Break down coding projects into structured tasks
• 🔄 Track project progress with weighted calculations
• 📊 View tasks organized by development phase
• ⏲️ Update task status as work progresses
• 📝 Automatically extract keywords from coding queries
• 📏 Estimate project complexity based on technologies
• 🗂️ Organize tasks by implementation phase

Quick Start

Option 1: Use the NPM Package

# Install globally
npm install -g mcp-task-manager

# Run the server
mcp-task-manager

Option 2: Use the Hosted Version

Add this to your .cursor/mcp.json file:

{
  "mcpServers": {
    "task-manager": {
      "url": "https://gvb-8o3fdody2-nbatans-projects.vercel.app/sse"
    }
  }
}

Option 3: Run It Directly

# Clone the repository
git clone https://github.com/yourusername/mcp-task-manager.git
cd mcp-task-manager

# Install dependencies
npm install

# Start the server
npm start

Using with Cursor

Once the MCP Task Manager is running, you can use it in Cursor to:

1. Process coding queries into tasks:

   Please break down this project: "Create a React dashboard with authentication and dark mode"
   

2. View project tasks:

   Show me the tasks for the dashboard project
   

3. Update task status:

   Mark the "Project Setup" task as completed
   

4. List all projects:

   Show me all my coding projects
   

API Endpoints

The MCP Task Manager exposes these endpoints:

Method	Endpoint	Description
GET	/api/tasks	Get all tasks
GET	/api/tasks/:id	Get a task by ID
POST	/api/tasks	Create a new task
PUT	/api/tasks/:id	Update a task
DELETE	/api/tasks/:id	Delete a task
POST	/api/ai/process-query	Process a coding query into tasks
PUT	/api/ai/tasks/status	Update task status
GET	/api/ai/projects	Get all coding projects
GET	/api/ai/projects/:projectId	Get tasks for a specific project
GET	/sse	MCP Server-Sent Events endpoint

Hosting Options

Deploy to Vercel

The easiest way to host your own instance is with Vercel:

1. Fork this repository
2. Import it in Vercel
3. Deploy

Self-Hosted Server

For a dedicated server:

# Install PM2 for process management
npm install -g pm2

# Start the server with PM2
pm2 start src/index.js --name "mcp-task-manager"
pm2 startup
pm2 save

Configuration Options

Environment Variables

• PORT: The port to run the server on (default: 3000)
• VERCEL: Set to '1' when running on Vercel

Development

# Install dependencies
npm install

# Run in development mode
npm run dev

How It Works

This MCP server implements the Model-Context-Protocol pattern:

• Models: Handle task data storage and retrieval
• Context: Represent the state of requests and tool calls
• Protocols: Define the business logic for each operation

The architecture follows a clean, modular design that makes it easy to extend.

License

MIT

Credits

Created by Your Name