MCP Server

A project management API and MCP server built with Flask and async Python. Includes endpoints for projects, tasks, and dependencies, and an MCP server for tool-based automation.

Features

• REST API for projects, tasks, and dependencies
• Marshmallow schemas for serialization/validation
• SQLite database by default (configurable)
• Async MCP server for tool automation
• Postman collection for API testing

Setup Instructions

1. Clone the repository

git clone <your-repo-url>
cd mcp-server

2. Create and activate a virtual environment (recommended)

python -m venv .venv
.venv\Scripts\activate  # On Windows
# Or
source .venv/bin/activate  # On macOS/Linux

3. Install dependencies

pip install -r requirements.txt

4. Database setup

• By default, uses SQLite (instance/default.db).
• To use another database, set the DATABASE_URL environment variable in your shell or .env file.
• Example for PostgreSQL:

  set DATABASE_URL=postgresql://user:password@localhost/dbname  # Windows
  export DATABASE_URL=postgresql://user:password@localhost/dbname  # macOS/Linux

5. Running the Flask API server

python run_flask.py

• The API will be available at http://127.0.0.1:5000/api/

6. Running the MCP server

python run_mcp.py

7. API Testing

• Import the provided Postman collection (MCP-Server-API.postman_collection.json) into Postman.
• Set the base_url variable to http://127.0.0.1:5000/api.

8. Environment Variables

• DATABASE_URL: Database connection string (default: SQLite)
• SECRET_KEY: Flask secret key
• DEBUG: Set to True for debug mode

---

Project Structure

• app/ - Flask app, models, routes, schemas
• mcp_server/ - MCP server and tool definitions
• instance/ - Database file (default.db)
• run_flask.py - Run Flask API
• run_mcp.py - Run MCP server
• requirements.txt - Python dependencies
• README.md - This file

License

MIT