anup-shesh/garmin-local-mcp

What it does

Garmin Local MCP is an MCP server that creates a personal data warehouse for Garmin fitness and health data. It provides incremental synchronization of your Garmin data into a SQLite database you control locally, eliminating dependency on cloud services. The tool processes raw data into meaningful insights including serverside trends, lag-aware correlations, personal baselines, and automated anomaly detection.

Key features

• Incremental sync into local SQLite database
• Serverside trend analysis and baseline calculations
• Lag-aware correlation detection for health metrics
• Anomaly detection for identifying unusual patterns
• Offline-first architecture - analysis works when Garmin API is down
• Compact table output format for easy integration
• Cross-platform support (Windows, macOS, Linux)

How to set up

Install via Python package manager, configure Garmin API credentials, and initialize the local SQLite database. The MCP server integrates with Claude or other compatible AI tools to provide natural language queries over your health data. Full setup documentation is available on the GitHub repository.