Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Repository cleanup: Removed old MCP server files (CLAUDE_DESKTOP.md, mcp_server.py, requirements.txt). This repo is now focused exclusively on the Clawdbot skill for automated health monitoring. Claude Desktop users should use the dedicated MCP server at github.com/eversonl/garmin-health-mcp-server. Both can coexist with shared authentication.
What This Skill Does
Connects to Garmin Connect and retrieves health metrics like sleep stages, Body Battery, HRV, heart rate, stress, and workout data. Supports natural language queries against your personal data and generates interactive HTML dashboards with Chart.js. Works with FIT/GPX files for route and elevation analysis.
Direct Garmin Connect integration with natural language querying avoids manual CSV exports or navigating the Garmin app for every data lookup.
When to use it
- Checking last night's sleep score and stage breakdown
- Reviewing weekly Body Battery trends before a hard training day
- Listing all workouts from the past month with pace and calories
- Asking why resting heart rate spiked on a specific day
- Generating a 90-day HRV trend chart to track recovery over time
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: how was my recovery this week?
- 1Run garmin_data.py body_battery --days 7 to fetch Body Battery readings
- 2Parse JSON output for daily values and calculate weekly average
- 3Compare daily highs and lows to identify recovery trend
- 4Cross-reference with garmin_data.py activities --days 7 to check training load
- 5Summarize average Body Battery, trend direction, and flag any low-recovery days
Weekly average Body Battery was 62, trending down from 74 on Monday to 48 by Friday, likely due to back-to-back training sessions on Wednesday and Thursday.
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.