Coach training and diet from wearable data
The poster uses Claude with Apple Health, nutrition and lifting app data, plus medication and bloodwork files, and has it maintain an HTML dashboard. Task: Suggested implementation: Integrate wearable data (e.g., Garmin) into a time-series database like InfluxDB. Configure an automated agent (e.g., Claude Code) to run daily, analyzing metrics such as sleep, stress, and training load. The agent calculates a readiness score and adjusts the training plan, while a human reviews and approves the final output before execution.; tools: Claude. Human role: Source evidence is unavailable regarding specific human roles in the provided text.; limitations: Prerequisites include server infrastructure for data polling and scheduled tasks. Integration uncertainty exists regarding API stability and data normalization. Human approval is mandatory before publishing or acting on any generated plans. Source reliability is low due to missing original content..
Tools
Published on Agent Work Atlas:
Overview
The poster uses Claude with Apple Health, nutrition and lifting app data, plus medication and bloodwork files, and has it maintain an HTML dashboard.
Workflow
Suggested implementation: Integrate wearable data (e.g., Garmin) into a time-series database like InfluxDB. Configure an automated agent (e.g., Claude Code) to run daily, analyzing metrics such as sleep, stress, and training load. The agent calculates a readiness score and adjusts the training plan, while a human reviews and approves the final output before execution.
Reported result
Actual outcomes are unknown as the source is inaccessible (HTTP 403). Proposed acceptance check: Verify that the generated training plans align with subjective fatigue levels and that no overtraining incidents occur over a 4-week pilot period.
Human role
Source evidence is unavailable regarding specific human roles in the provided text.
Limitations
Prerequisites include server infrastructure for data polling and scheduled tasks. Integration uncertainty exists regarding API stability and data normalization. Human approval is mandatory before publishing or acting on any generated plans. Source reliability is low due to missing original content.
Try it
Act as a fitness data analyst. I have exported my last 30 days of Garmin data to [replaceable_input: CSV_file_path]. Please analyze sleep, stress, and training load metrics. Calculate a daily training-readiness score (0-100). Based on this, draft a 7-day training plan that avoids overtraining. Output the results in a JSON format with keys: 'daily_scores', 'weekly_plan', and 'risk_notes'. Before finalizing, ask me to confirm my subjective energy levels for the last 3 days to calibrate the score. Do not publish or send this plan; just provide the draft for my review.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.