AI Task Management Assistant
A Model Context Protocol (MCP) server prototype that provides task management capabilities for AI assistants. This server demonstrates how Claude and other MCP-compatible AI models can access and query task data stored in Supabase.
This is in an early experimental prototype primarily designed for:
- Learning about MCP server implementation
- Prototyping AI-database interactions
- Testing the Model Context Protocol capabilities
Features
- Retrieve overdue tasks from a Supabase database
- Designed for integration with Claude and other MCP-compatible AI assistants
Prerequisites
- Node.js 18+
- npm or yarn
- A Supabase account with a project set up
- A task database with the required schema
Installation
-
Clone this repository:
git clone <your-repo-url> cd ai-tasks-mcp -
Install dependencies:
npm install -
Set up environment variables by creating a
.envfile with the following content:NEXT_PUBLIC_SUPABASE_URL=your_supabase_url NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
Example query: "What tasks do I have that are overdue?"
Testing with Claude
To test this MCP server with Claude:
. Use Claude Web, Desktop, or API with MCP capabilities enabled . Ask Claude to connect to your local MCP server
# claude code example
npm install -g @anthropic-ai/claude-code
claude mcp add "project-task-manager" npx txs index.ts
claude mcp list
4. Prompt for overdue tasks:
> "Which tasks are currently overdue?"
5. The tool will return a formatted list of overdue tasks direct from the database.