ExpenseTracker MCP Server
A production-ready Model Context Protocol (MCP) server that turns your AI assistant into a persistent personal finance manager using PostgreSQL.
This project allows Claude Desktop or any MCP-enabled agent to securely store, analyze, and export expense data using real database transactions instead of chat memory.
What Is It?
ExpenseTracker MCP is a local backend service exposing structured financial tools to AI agents.
It enables your assistant to:
- Store expenses permanently
- Edit and delete past records
- Summarize spending patterns
- Detect top spending categories
- Generate monthly reports
- Export data for accounting or tax use
How It Works
- Claude Desktop sends a tool request using MCP.
- ExpenseTracker MCP receives the request.
- The server executes the database operation in PostgreSQL.
- Results are returned to Claude as structured JSON.
- Logs are written to
expense_tracker.log.
The AI never invents data — it only queries your real database.
How To Run Using uv
Install dependencies
uv add fastmcp psycopg2-binary python-dotenv
Create .env file
DB_HOST=localhost
DB_PORT=5432
DB_NAME=expense_tracker
DB_USER=expense_user
DB_PASSWORD=your_password
Start MCP server
For Testing
uv run fastmcp dev main.py
For Run
uv run fastmcp run main.py
How to connect to Claude Desktop
uv run fastmcp install claude-desktop main.py
Restart Claude Desktop
Contribution
Contributions are welcome.
- Fork the repository
- Create a feature branch
- Add or improve MCP tools or documentation
- Submit a pull request with a clear description of your changes