Chat with your Garmin Connect data using AI to gain deep insights into your training, performance trends, and recovery metrics.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install garmin-skill
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install garmin-skill using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Garmin Skill?
The Garmin Skill allows AI agents to interface directly with Garmin Connect data through the Transition platform. By integrating this skill into Openclaw Skills, developers and athletes can transform raw activity logs into a conversational intelligence layer. It provides a bridge between wearable hardware and large language models, allowing users to ask complex questions about their physiology, training history, and performance trends in plain English.
This skill is powered by Transition, which handles the synchronization of Garmin data and provides a specialized AI Coach endpoint. Whether you are tracking VO2 Max trends, analyzing heart rate zones, or calculating your Training Stress Balance (TSB), this skill provides the necessary endpoints to build sophisticated fitness assistants and automated coaching workflows.
Garmin Skill Use Cases
- Analyze monthly running or cycling volume trends compared to previous periods.
- Monitor recovery by checking TSB (Training Stress Balance) and fatigue levels to avoid overtraining.
- Generate structured workouts (WOD) for running, swimming, or strength training on the fly.
- Push scheduled workouts directly to a Garmin device for immediate execution.
- Query specific performance metrics like FTP, threshold paces, or heart rate zones using natural language.
How Garmin Skill Works
- The AI agent authenticates with the Transition API using a secure X-API-Key passed in the request header.
- When a user asks a question, the agent sends the message to the AI Coach endpoint which has full context of the user's Garmin history.
- For quantitative analysis, the agent retrieves structured data from endpoints like PMC (Performance Management Chart) or Performance Stats.
- The agent processes the CTL (Fitness), ATL (Fatigue), and TSB (Form) values to provide scientific training advice.
- If a workout is agreed upon, the agent can trigger a POST request to push that specific workout directly to the user's connected Garmin hardware.
Garmin Skill Setup
To get started with this skill in Openclaw Skills, follow these steps:
- Download the Transition application and link your Garmin Connect account.
- Navigate to Settings > API Keys within the Transition app and select Generate New Key.
- Set your environment variable to authenticate your requests:
export TRANSITION_API_KEY="tr_live_xxxxxxxxxxxxxxxxxxxxx"
- For unauthenticated testing of the Workout of the Day (WOD) feature, you can use:
curl "https://api.transition.fun/api/v1/wod?sport=run&duration=45"
Garmin Skill Data Schema & Taxonomy
The Garmin Skill organizes data into several logical domains to help Openclaw Skills agents parse information efficiently:
| Domain | Data Point | Description |
|---|---|---|
| Performance | PMC | CTL (Fitness), ATL (Fatigue), and TSB (Form) calculations. |
| Metrics | Stats | FTP, threshold paces, and personalized heart rate zones. |
| Planning | Workouts | Scheduled activities including duration, sport type, and intensity. |
| Profile | Athlete | Basic user metadata and sync status from Garmin Connect. |
| Intelligence | Coach | Contextualized AI chat history and training load analysis. |
Garmin Skill Advanced Features
- Scientific Fatigue Tracking: Automatically calculate TSB to warn users if they are at risk of injury before recommending hard efforts.
- Bi-directional Sync: Not only reads data but can write structured workouts directly to Garmin wearable devices.
- Zero-Auth WOD: Access a public endpoint for generating random structured workouts without requiring a full user account.
- Performance Trend Analysis: Compare historical periods (e.g., month-over-month) to visualize fitness progression via the AI Coach interface.
- Flexible Rate Limiting: Support for both free tier testing and high-throughput paid tiers for professional coaching platforms.
SKILL.md
Loading
Related Openclaw Skills
Apple Health Skill
An AI-driven integration that enables users to query their Apple Health data using natural language for personalized fitness coaching and performance tracking.
MeetGeek
A terminal-based interface to query MeetGeek for meeting summaries, transcripts, and searchable intelligence.
X Articles
A high-performance AI writing and automation skill designed to create and publish viral long-form articles directly to X.
X Algorithm Mastery
A comprehensive technical guide and rule-set for AI agents to master X (Twitter) engagement and content strategy.
Race Finder
A powerful tool for AI agents to search and retrieve real-time endurance race data including running, triathlons, and cycling events.
Strava AI Skill
A conversational AI integration for Strava that enables athletes to query performance trends, analyze fatigue, and generate structured workouts.
Featured*
