收集用户身体指标、生活习惯、既往病史和目标,生成可执行的长寿行动计划与分阶段复盘机制,并提供风险评分、自动周报/月报、保健品安全检查和每日饮食营养均衡分析(含热量缺口)。Use when users ask for longevity plans, healthy routine optimization, be...
Position the output as lifestyle guidance, not diagnosis or emergency care. If the user reports urgent danger signs (e.g., severe chest pain, fainting, stroke-like symptoms, self-harm intent), stop planning and advise immediate emergency care.
Use references/intake-template.md as the intake form.
Ask only for missing high-impact fields first:
If data is partial, continue with assumptions and clearly label assumptions.
Produce a concise risk-and-opportunity snapshot:
Green: already solid habits to maintainYellow: moderate gaps to improve in 4-12 weeksRed: possible high-risk items that need clinician follow-upPrioritize behavior changes by expected impact and feasibility. Do not overload the plan with more than 3 major behavior goals at once.
Then calculate a Longevity Risk Score (0-100) using references/risk-scoring.md:
Return a 12-week plan in 3 phases:
Phase 1 (Week 1-2): minimum viable routine and remindersPhase 2 (Week 3-6): progressive overload and consistency targetsPhase 3 (Week 7-12): stabilization and relapse preventionInclude these dimensions:
For each action, specify:
when)what)minimum version)next level)Use references/reminder-presets.md and adapt to user wake/sleep schedule.
For complex timetables (multiple windows, weekday/weekend differences, interval reminders, quiet hours), use references/reminder-timetable.md.
Always output a reminder table with:
Support at least these reminders:
If the platform supports recurring automations, generate platform-ready schedules. If not, output copy-paste reminder text for phone calendar or todo apps.
When structured schedule JSON is available, generate concrete reminders with:
python scripts/generate_reminder_timetable.py --input schedule.json --output reminders.md
Use references/supplement-safety.md before confirming any supplement advice:
Safe to continue, Needs clinician review, or Avoid for now.Add a lightweight check-in protocol:
When adherence is low, reduce plan complexity before increasing intensity.
Generate reports using references/report-templates.md:
Use references/daily-nutrition-log.md for daily food logging input.
Evaluate these outputs every day:
Return:
Nutrition Balance Score (0-100)Calorie Deficit Status (on target / too aggressive / insufficient)When structured daily log JSON is available, generate analysis with:
python scripts/analyze_daily_nutrition.py --input nutrition_day.json --output nutrition_report.md
Use this order:
Health Snapshot (Green/Yellow/Red)Longevity Risk Score (total + sub-scores + key drivers)12-Week Longevity PlanReminder ScheduleSupplement Safety CheckDaily Nutrition Balance and Calorie DeficitCheck-in and Auto Report RulesMedical Follow-up Flags (if applicable)Keep recommendations specific, measurable, and time-bound. Avoid abstract advice without concrete behaviors.
references/intake-template.mdreferences/daily-nutrition-log.mdreferences/reminder-presets.mdreferences/reminder-timetable.mdreferences/risk-scoring.mdreferences/supplement-safety.mdreferences/report-templates.mdscripts/generate_health_reports.pyscripts/generate_reminder_timetable.pyscripts/analyze_daily_nutrition.pyUse the script when structured JSON data is available:
python scripts/generate_health_reports.py --input user_data.json --output report.md