A simple, powerful Todo list manager for Claude Desktop and other MCP-compatible AI assistants. Organize your tasks across different projects with priorities and never lose track of what needs to be done!
- π Simple Task Management - Create, complete, and archive tasks
- π Project Organization - Separate tasks by lists (perfect for different codebases/projects)
- π― Priority System - Low, Medium, High, and Urgent priorities
- π Time Tracking - Automatic timestamps for creation, completion, and archiving
- π Smart Sorting - Tasks sorted by priority and age automatically
- πΎ Persistent Storage - Uses PostgreSQL for reliable data storage
- Download DBngin - A simple database manager for Mac
- Install and open DBngin
- Click "+" and select PostgreSQL
- Start the PostgreSQL server
- Create two databases:
todo_mcp
(for your tasks)todo_mcp_test
(for testing)
- Create a
docker-compose.yml
file:
version: '3.8'
services:
postgres:
image: postgres:15
environment:
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
POSTGRES_DB: todo_mcp
ports:
- "5432:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
volumes:
postgres_data:
- Run:
docker-compose up -d
- Create test database:
docker exec -it <container_name> createdb -U postgres todo_mcp_test
Add this to your Claude Desktop MCP settings:
{
"mcpServers": {
"todo-mcp": {
"command": "npx",
"args": ["-y", "@floriscornel/todo-mcp@latest"],
"env": {
"DATABASE_URL": "postgresql://postgres:postgres@localhost:5432/todo_mcp"
}
}
}
}
Ask Claude things like:
- "Create a new list called 'Frontend Project'"
- "Add a task to fix the login bug with high priority"
- "Show me all my urgent tasks"
- "Complete the task about updating dependencies"
You: "I'm reviewing this PR and found several issues. Can you help me track them?"
Claude: "I'll help you create tasks for the issues you found. First, let me create a list for this PR."
You: "Create a list called 'PR-123 Review Issues'"
Claude: *Creates the list*
You: "Add these tasks:
- Fix memory leak in UserService (high priority)
- Add error handling for API calls (medium)
- Update outdated dependencies (low)
- Add unit tests for new features (high)"
Claude: *Creates all tasks with appropriate priorities*
You: "Show me the high priority items first"
You: "I just got 5 bug reports from users. Help me organize them by priority."
Claude: "Let me create a 'Bug Reports - Sprint 24' list and add these issues. What are the bugs?"
You: "Login fails on mobile Safari, checkout process crashes, slow page load, typo in footer, and dark mode toggle broken"
Claude: *Creates tasks and asks about severity*
Claude: "I've added these. The login and checkout issues sound urgent - should I mark them as high priority?"
You: "Yes, and show me all urgent tasks across all my lists so I can plan my day"
You: "I'm starting work on a new dashboard feature. Help me break it down into tasks."
Claude: "Great! Let me create a 'Dashboard Feature' list. What components do you need to build?"
You: "I need to design the layout, create API endpoints, build the frontend components, write tests, and update documentation"
Claude: *Creates structured tasks*
Claude: "I've created tasks for each component. Should I mark the API endpoints as high priority since the frontend depends on them?"
You: "Perfect! Also add a task to review the design with the team before I start coding"
You: "I'm learning React Native. Can you help me track my progress?"
Claude: "I'll create a 'React Native Learning' list. What topics do you want to cover?"
You: "Navigation, state management, API integration, testing, and deployment"
Claude: *Creates learning tasks*
You: "I just finished the navigation tutorial. Mark that as complete and show me what's next"
Claude: "Great progress! Next up is state management. I can also help you find resources for that topic."
You: "I need to refactor this legacy codebase. There's so much to do, I don't know where to start."
Claude: "Let's break it down systematically. I'll create a 'Legacy Refactor' list. What are the main areas that need work?"
You: "Database queries are slow, no error handling, outdated dependencies, no tests, and the UI is a mess"
Claude: *Creates organized tasks with priorities*
Claude: "I've prioritized database performance and error handling as urgent since they affect users directly. Want me to help you tackle the database issues first?"
You: "What did I work on yesterday and what should I focus on today?"
Claude: "Let me check your completed tasks from yesterday and show you today's priorities."
*Shows completed tasks and upcoming high-priority items*
Claude: "Yesterday you completed the user authentication fix and API documentation. Today's urgent tasks are the payment gateway integration and the mobile responsive fixes."
You: "Perfect! Add a new task to review the payment gateway with the security team before deploying"
You: "Help me plan this 2-week sprint. I have tasks scattered across different lists."
Claude: "I'll show you all your tasks grouped by priority. Which lists should I include in the sprint planning?"
You: "Include 'Frontend Bugs', 'API Development', and 'User Stories - Q4'"
Claude: *Aggregates and prioritizes tasks*
Claude: "Here are your urgent and high-priority tasks. Based on your usual velocity, I'd recommend focusing on these 8 tasks for the sprint. Should I create a 'Sprint 12' list to track them?"
You: "This function is behaving weirdly. Help me track what I need to investigate."
Claude: "I'll create an 'Investigation - UserAuth Bug' list. What symptoms are you seeing?"
You: "Users can't log in on weekends, but weekdays work fine. It's bizarre."
Claude: "That is strange! Let me add investigation tasks:
- Check server logs for weekend patterns
- Review cron jobs and scheduled tasks
- Test with different user accounts
- Check database connection pooling
- Review load balancer configuration"
You: "Great breakdown! I just checked the logs - mark that complete and add a task to check if our SSL certificate renewal happens on weekends"
- Persistent Context: Your tasks persist across conversations
- Natural Language: "Show me what I was working on last week"
- Smart Prioritization: Claude can help reprioritize based on deadlines
- Code-Aware Tasks: "Add a task to refactor this function" (while viewing code)
- Context Switching: Quickly switch between coding and task management
- Inline Planning: Plan refactoring while reviewing code
- Pick Up Where You Left Off: "What was I working on before the meeting?"
- Team Handoffs: "Show the new developer what needs to be done on this feature"
- Progress Tracking: "How much of the authentication system is complete?"
Think of lists as different projects or categories:
- Frontend Repo - Tasks for your React app
- Backend API - Server-side todos
- Personal - Non-work related tasks
Each task has:
- Name (5-120 characters) - Keep it concise and actionable
- Description (optional, up to 500 characters) - Add details if needed
- Priority - Low β Medium β High β Urgent
- Timestamps - When created, completed, or archived
- Tasks are automatically sorted by priority (urgent first) and age
- Completed tasks are kept for reference but don't clutter your active list
- Archive tasks when they're no longer relevant
The MCP provides these tools for Claude:
getLists()
- Show all your project listscreateList(name, description?)
- Create a new project listgetTasks(list, includeCompleted?)
- Get tasks from a specific listcreateTask(list, name, description?, priority?)
- Add a new taskcompleteTask(taskId)
- Mark a task as donearchiveTask(taskId)
- Archive a task (removes from active view)
# Main database (required)
DATABASE_URL="postgresql://username:password@localhost:5432/todo_mcp"
# Test database (optional, for development)
TEST_DATABASE_URL="postgresql://username:password@localhost:5432/todo_mcp_test"
# Local PostgreSQL
DATABASE_URL="postgresql://postgres:postgres@localhost:5432/todo_mcp"
# Docker
DATABASE_URL="postgresql://postgres:postgres@localhost:5432/todo_mcp"
# Remote database (replace with your details)
DATABASE_URL="postgresql://user:pass@your-db-host:5432/todo_mcp"
- Node.js 18+
- PostgreSQL 12+
- npm or yarn
# Clone the repository
git clone https://github.com/floriscornel/todo-mcp.git
cd todo-mcp
# Install dependencies
npm install
# Set up environment variables
cp .env.example .env
# Edit .env with your database URLs
# Run database migrations
npm run db:migrate
# Run tests
npm test
# Start development
npm run dev
npm test # Run tests once
npm run test:watch # Run tests in watch mode
npm run test:coverage # Run tests with coverage report
We welcome contributions! Here's how to get started:
- Fork the repository
- Create a feature branch:
git checkout -b feature/amazing-feature
- Make your changes and add tests
- Run the test suite:
npm test
- Commit your changes:
git commit -m 'Add amazing feature'
- Push to the branch:
git push origin feature/amazing-feature
- Open a Pull Request
- Write tests for new features
- Follow the existing code style (we use Biome for formatting)
- Update documentation for user-facing changes
- Keep commits focused and descriptive
This project is licensed under the MIT License - see the LICENSE file for details.
- Built with the Model Context Protocol by Anthropic
- Uses Drizzle ORM for database operations
- Inspired by the need for better task management in AI-assisted development
- Issues: GitHub Issues
- Discussions: GitHub Discussions
- Documentation: MCP Documentation
- Web interface for task management
- Task due dates and reminders
- Task dependencies and subtasks
- Export/import functionality
- Team collaboration features
- Integration with popular task managers
Made with β€οΈ for the AI-assisted development community