Health Copilot: Eating, Sleep, and Exercise Tracking

Use this skill as the central hub for personal health tracking, analysis, and Feishu reporting.

Overview

This skill is for practical personal-health management, not medical diagnosis.

It provides a unified layer for:

• Recognition: Analyzing food photos, sleep screenshots, and workout summaries.
• Persistence: Mapping and writing results into a structured Feishu Bitable.
• Reporting: Rebuilding daily/weekly summary records and monthly dashboards.

Role & Workflow

1. Classify: Identify if the request is about nutrition, sleep, exercise, or cross-domain reporting.
2. Analyze: Perform domain-specific extraction or analysis (conservative and evidence-based).
3. Persist: Map the results into the target Feishu schema (default English-only).
4. Automate: Run maintenance scripts for monthly/weekly summaries when requested.

Routing branches

1. Nutrition

• Trigger: log calories, nutrition analysis, meal photos, bento photos.
• Reference: references/nutrition.md.
• Behavior: Analyze photo -> Estimate nutrients -> Upsert to Nutrition Meals table.

2. Exercise

• Trigger: log workout, exercise analysis, app screenshots (Garmin, Strava, Apple Health).
• Reference: references/exercise.md.
• Behavior: Analyze screenshot -> Extract metrics -> Assess load -> Upsert to Workouts table.

3. Sleep

• Trigger: log sleep, sleep analysis, recovery status, app screenshots (AutoSleep, Oura, etc.).
• Reference: references/sleep.md.
• Behavior: Analyze screenshot -> Extract stages/HRV -> Upsert to Sleep Recovery table.

4. Cross-domain / Reporting

• Trigger: weekly health review, rebuild calendar, refresh dashboard, monthly summary.
• Reference: references/cross-domain.md.
• Behavior: Rebuild Monthly Health Calendar -> Rebuild Weekly Health Assessment -> Refresh Dashboard.

Universal rules

• Internal Schema: Always target the English-only internal schema for persistence to ensure script reliability.
• Config-Driven: Use references/config-template.md and scripts/config_loader.py for all environment variables.
• Conservative Analysis: Extract only what is visible; do not fake precision.
• Separation: Keep analysis facts separate from interpretive suggestions.

Output rule

Default response shape:

1. Conclusion (what was analyzed/updated)
2. Evidence (the data points)
3. Uncertainty (missing info)
4. Next step (action or report link)

Script mapping

• scripts/bootstrap_health_tables.js: Create tables/fields.
• scripts/rebuild_monthly_calendar.py: Aggregate daily rows.
• scripts/rebuild_weekly_assessment.py: Generate weekly reports.
• scripts/build_monthly_dashboard.js: Create/refresh dashboards.

Installation

clawdhub install personal-health-router