Get a daily training plan from your own watch data
Idea + starter prompt, not run by us yet
For: Runners and gym-goers who already track sleep, load and stress on a wearable
One commenter pulls Garmin data into a database on their own server every hour. A scheduled daily Claude Code task reads load, sleep and stress, scores training readiness, adjusts the plan and publishes it to a small web page for their phone. They also loaded years of past activity files for context.
Steps
- Check that your watch or app can export data, and decide where it will live. Keep it on a machine you control.
- Start with a one-off export and ask Claude to describe what it sees before you automate anything.
- Ask Claude to explain how it scores readiness so you can judge if the logic makes sense.
- Schedule the daily run only after a week of manual runs that you agree with.
- Treat the plan as a suggestion. See a doctor for pain or health issues, and do not use it to adjust medication.
Starter prompt
Trial first: work on a copy of my export, cut down to the last [7] days, and save everything to [trial-plan.md]. Never edit the original file. Here is my training data: [file], covering [sleep, load, stress, workouts] for [period]. My goal is [goal, e.g. a 10 km race in 3 months]. First describe what you see in my history and point out gaps. Then propose how you would score my training readiness each morning and show me the formula. Stop and wait for my yes on the formula before you draft a one-week plan. Explain your reasoning for each day. The plan is a suggestion that only reads and drafts, so change no device or app setting. Give no medical advice, and tell me to see a professional if the data suggests a health problem.
Original post: Reddit · found via agentusecases.ai. Written up in our own words. We haven't run it, so test it on made-up data first.