Todo MCP Server
A Todo and Task Manager server for creating, updating, and managing tasks, using a SQLite database.
Todo MCP Server
A comprehensive Todo/Task Manager MCP (Model Context Protocol) Server built with Node.js, TypeScript, and SQLite. This server provides tools, resources, and prompts for managing todos through Claude and other MCP-compatible clients.
Features
š ļø Tools (7 available)
- create_todo - Create new todos with title, description, priority, due date, and tags
- update_todo - Update existing todos (any field)
- delete_todo - Delete todos by ID
- complete_todo - Mark todos as completed
- list_todos - List todos with filtering (completion status, priority, tag, limit)
- search_todos - Search todos by title, description, or tags
- get_todo_stats - Get comprehensive todo statistics
š Resources (9 available)
- todos://all - All todos (JSON)
- todos://completed - Completed todos (JSON)
- todos://pending - Pending todos (JSON)
- todos://overdue - Overdue todos (JSON)
- todos://todo/{id} - Specific todo by ID (JSON)
- todos://priority/{priority} - Todos by priority (JSON) with autocomplete
- todos://tag/{tag} - Todos by tag (JSON) with autocomplete
- todos://stats - Todo statistics (JSON)
- todos://tags - All available tags (JSON)
š¬ Prompts (5 available)
- create_todo_prompt - Interactive todo creation
- daily_review - Daily todo review and planning
- weekly_planning - Weekly planning session
- productivity_analysis - Analyze productivity patterns
- task_breakdown - Break down complex tasks
Installation
-
Clone or create the project directory:
mkdir todo-mcp-server cd todo-mcp-server -
Install dependencies:
npm install -
Build the project:
npm run build
Usage
Basic Server Operations
-
Test the todo service:
npm run test -
Run the MCP server:
npm start -
Development mode (with file watching):
npm run dev
Testing with MCP Inspector
-
Start the inspector:
npx @modelcontextprotocol/inspector node dist/index.js -
Open the web interface:
- Follow the URL provided in the terminal output
- Use the provided authentication token
Claude Desktop Integration
Add this configuration to your Claude Desktop config file:
{
"mcpServers": {
"todo-manager": {
"command": "node",
"args": ["/path/to/todo-mcp-server/dist/index.js"],
"env": {
"TODO_DB_PATH": "/path/to/your/todos.db"
}
}
}
}
Configuration
Environment Variables
- TODO_DB_PATH: Path to SQLite database file (default:
todos.dbin current directory)
Database
The server uses SQLite with better-sqlite3 for data persistence. The database is automatically created with the following schema:
- todos table with fields: id, title, description, completed, priority, due_date, tags, created_at, updated_at
- Indexes on completion status, priority, due date, and creation date for optimal performance
API Examples
Creating a Todo
// Using the create_todo tool
{
"title": "Review documentation",
"description": "Read through the MCP protocol documentation",
"priority": "high",
"dueDate": "2024-06-27T10:00:00.000Z",
"tags": ["work", "learning"]
}
Filtering Todos
// Using the list_todos tool
{
"completed": false,
"priority": "high",
"limit": 10
}
Resource Access
GET todos://pending # All pending todos
GET todos://priority/high # High priority todos
GET todos://tag/work # Todos tagged with "work"
GET todos://stats # Todo statistics
Architecture
src/
āāā database.ts # SQLite database layer with better-sqlite3
āāā index.ts # Main server entry point
āāā mcp-server.ts # MCP server implementation
āāā todo-service.ts # Business logic layer
āāā test.ts # Test suite
āāā types.ts # TypeScript type definitions
Key Components
- TodoDatabase - Direct SQLite interface with better-sqlite3
- TodoService - Business logic with validation and error handling
- TodoMcpServer - MCP protocol implementation with tools, resources, and prompts
Development
Scripts
npm run build- Compile TypeScript to JavaScriptnpm run dev- Development mode with file watchingnpm start- Start the production servernpm run test- Run the test suite
Technology Stack
- Runtime: Node.js
- Language: TypeScript
- Database: SQLite with better-sqlite3
- Protocol: Model Context Protocol (MCP)
- Validation: Zod schemas
Testing
The project includes comprehensive testing:
npm run test
This will:
- Create an in-memory database
- Test all CRUD operations
- Verify search and filtering functionality
- Check statistics generation
- Validate tag management
Troubleshooting
Common Issues
- Database permission errors: Ensure the directory is writable
- Port conflicts: The MCP server uses stdio, no ports needed
- TypeScript errors: Run
npm run buildto check for compilation issues
Debugging
Enable debug logging by setting:
export DEBUG=todo-mcp-server
License
MIT License - see LICENSE file for details.
Contributing
- Fork the repository
- Create a feature branch
- Add tests for new functionality
- Ensure all tests pass
- Submit a pull request
Changelog
v1.0.0
- Initial release
- Complete MCP implementation with tools, resources, and prompts
- SQLite database with better-sqlite3
- Comprehensive test suite
- Node.js compatibility (migrated from Bun)
Related Servers
Geekbot MCP
An MCP server that integrates Geekbot data and tools with AI models.
Asana
Interact with the Asana API to manage tasks, projects, and workspaces.
Jean Memory
Premium memory consistent across all AI applications.
Browser Use MCP Server
Automate browser actions using natural language commands. Powered by Playwright and supports multiple LLM providers.
Logseq
Control and interact with a local Logseq graph for knowledge management and note-taking.
Lenny's Podcast Transcripts
Search 286 episodes of product management wisdom from Lenny Rachitsky. Semantic search across 300+ hours of transcripts.
waitlister-mcp
MCP server for the Waitlister API. Manage your waitlist subscribers, track signups, and log views through AI assistants like Claude, Cursor, and Windsurf.
Google Calendar
Interact with Google Calendar APIs to manage events and calendars.
Leantime MCP Bridge
An MCP proxy bridge for the Leantime project management system, forwarding JSON-RPC messages with proper authentication.
Browser Control MCP
Enables AI agents to control a local browser, manage tabs, and conduct research via a browser extension.