Health & Fitness Agent

A comprehensive fitness companion that creates personalized workout plans and nutrition guidance while syncing with your wearable devices. It pulls data from Apple Health, Fitbit, or Garmin to understand your activity levels, sleep quality, and recovery status, then adapts your training program accordingly. The agent balances progressive overload with adequate recovery, adjusts for real-life schedule disruptions, and provides evidence-based nutrition recommendations aligned with your specific goals.

Supported Platforms & Integrations

Platform	Integration Type	Features
Apple Health	Data Sync	Reads heart rate, steps, sleep, workouts, and active calories for adaptive planning
Fitbit	Wearable Sync	Pulls daily readiness scores, sleep stages, and activity zones for recovery-aware scheduling
Garmin Connect	Training Data	Imports training load, VO2 max estimates, and body battery for endurance-focused plans
MyFitnessPal	Nutrition Tracking	Cross-references meal logs with macro targets and adjusts nutrition recommendations
Strava	Activity Import	Syncs run, ride, and swim data for endurance athletes who track on Strava
Strong App	Workout Logging	Imports weight training logs to track progressive overload and suggest adjustments

When to Use

• Starting a new fitness routine — builds a progressive plan matched to your current fitness level, available equipment, and schedule constraints
• Training for a specific event — creates periodized programs for races (5K to marathon), competitions, or physical challenges with proper tapering
• Breaking through a plateau — analyzes your training history to identify stagnation patterns and introduces targeted variation
• Recovering from injury — designs modified programs that work around limitations while maintaining fitness in unaffected areas
• Balancing multiple fitness goals — structures concurrent training for seemingly conflicting goals like building muscle while improving endurance
• Optimizing nutrition for performance — calculates macronutrient targets based on training load, body composition goals, and metabolic estimates

Alternatives to Consider

• Use a certified personal trainer if you have complex medical conditions, serious injuries, or need hands-on form correction
• Use dedicated apps (JEFIT, Nike Training Club) if you prefer video-guided workouts over text-based plans
• Use a registered dietitian if you have eating disorders, severe food allergies, or medical nutrition therapy needs

Quick Start

Copy
agent: health-fitness-agent
config:
  goals:
    primary: "muscle_gain"
    secondary: "general_fitness"
  experience_level: "intermediate"
  weekly_sessions: 4
  session_duration: 60
  equipment: ["full_gym"]
  health_data_source: "apple_health"
  dietary_preferences: ["no_dairy"]
  current_stats:
    age: 32
    weight_kg: 78
    height_cm: 178

Example Prompt

I want to build muscle while staying relatively lean. I'm 32, 
178cm, 78kg, intermediate lifter (2 years). I have access to a 
full gym 4 days per week, 60 minutes per session. I'm lactose 
intolerant. Use my Apple Health data for recovery adjustments.

Example Output

## Training Program: Lean Muscle Gain
Split: Upper/Lower (4 days) | Phase 1 of 3 | Weeks 1-4

### Monday — Upper A (Push Focus)
1. Barbell Bench Press: 4x6-8 @ RPE 8
2. Overhead Press: 3x8-10 @ RPE 7
3. Incline Dumbbell Press: 3x10-12
4. Cable Lateral Raise: 3x12-15
5. Tricep Pushdown: 3x10-12
6. Face Pull: 3x15-20
Estimated duration: 55 minutes

### Nutrition Targets (Training Days)
Calories: 2,650 | Protein: 175g | Carbs: 310g | Fat: 78g
Dairy-free protein sources: chicken, eggs, soy milk, pea protein...

Advanced Configuration

Platform-Specific Settings

Copy
apple_health:
  sync_metrics:
    - resting_heart_rate
    - hrv
    - sleep_hours
    - active_energy
  recovery_threshold:
    hrv_below_baseline: 15  # percent drop triggers rest day
    sleep_below_hours: 6

nutrition:
  meal_timing: true
  pre_workout_window: 90  # minutes before training
  post_workout_window: 60
  hydration_reminders: true

Parameters Reference

Parameter	Type	Default	Description
goals.primary	string	required	Main goal: muscle_gain, fat_loss, endurance, strength, general_fitness
goals.secondary	string	null	Secondary goal to balance alongside primary
experience_level	string	"beginner"	Training experience: beginner, intermediate, advanced
weekly_sessions	number	3	Available training sessions per week (2-7)
session_duration	number	60	Available minutes per session
equipment	array	["bodyweight"]	Available equipment: full_gym, home_dumbbells, bodyweight, resistance_bands, kettlebell
health_data_source	string	null	Wearable platform: apple_health, fitbit, garmin, none
dietary_preferences	array	[]	Restrictions: vegetarian, vegan, no_dairy, no_gluten, keto, halal
deload_frequency	number	4	Weeks between deload/recovery weeks
progressive_overload	string	"auto"	Progression method: auto, linear, undulating, block
injury_notes	string	null	Current injuries or limitations to work around
include_cardio	boolean	true	Include cardio recommendations alongside resistance training

Core Concepts

Concept	Description
Progressive Overload	Systematically increasing training stimulus (weight, reps, or volume) over time to drive continuous adaptation
Recovery-Aware Scheduling	Uses HRV, sleep, and readiness data from wearables to determine if today should be a hard session, light session, or rest day
Periodization	Organizes training into phases (hypertrophy, strength, peaking, deload) to prevent plateaus and manage fatigue
Macro Cycling	Adjusts carbohydrate and calorie intake based on training vs. rest days to support performance and body composition simultaneously
RPE-Based Loading	Uses Rate of Perceived Exertion instead of fixed percentages, allowing natural daily strength fluctuations without derailing the program

┌──────────────┐     ┌───────────────┐     ┌──────────────┐
│  Wearable    │────▶│   Recovery    │────▶│   Session    │
│  Data Sync   │     │   Assessment  │     │   Planning   │
└──────────────┘     └───────────────┘     └──────┬───────┘
                                                   │
┌──────────────┐     ┌───────────────┐             │
│  Nutrition   │◀───│   Progress    │◀─────────────┘
│  Adjustment  │     │   Tracking    │
└──────────────┘     └───────────────┘

Workflow Examples

Scenario 1: Adapting to Poor Recovery

Input:

Apple Health shows my HRV dropped 20% this week and I only 
averaged 5.5 hours of sleep. I have a heavy leg day scheduled 
for today.

Output: The agent recommends converting today's heavy leg session to a moderate-intensity mobility and hypertrophy session. Squats change from 4x5 heavy to 3x10 moderate, a 15-minute mobility circuit is added, and tomorrow's rest day swaps with Thursday's session to allow recovery. Nutrition targets add 200 extra calories focused on carbs to support recovery.

Scenario 2: Travel Week with Limited Equipment

Input:

I'm traveling for work next week. Hotel gym only has dumbbells 
up to 25kg and a cable machine. How do I keep my program on track?

Output: A modified 3-day program using available equipment. Compound movements switch to dumbbell variants (DB bench, goblet squats, DB rows), rep ranges increase to compensate for lighter loads, and tempo manipulation (3-second eccentrics) is added to maintain stimulus. The agent marks the week as a strategic deload if it falls near the planned deload period.

Scenario 3: Cutting Phase Transition

Input:

I've been gaining for 4 months and want to start cutting. 
Currently 83kg at roughly 18% body fat. Target: 12-14%.

Output: A phased cutting plan: calories reduced by 300 per day (phase 1), training volume maintained but intensity slightly reduced, cardio added as two 30-minute LISS sessions, protein increased to 2.2g/kg to preserve muscle. Timeline estimate of 12-16 weeks with monthly check-ins for rate-of-loss adjustment. Includes refeed day protocol for weekends.

Best Practices

1. Sync your wearable data before each session — The recovery-aware scheduling only works with current data. Sync your watch or band before asking for today's recommendation to get the most accurate adjustment.

2. Log actual weights and reps, not just completion — Progressive overload tracking requires knowing what you actually lifted, not just that you did the workout. Accurate logging enables precise progression recommendations.

3. Do not skip deload weeks — They feel unproductive but are where adaptation actually consolidates. The agent schedules them strategically; trust the periodization even when you feel strong.

4. Track nutrition consistently for at least 2 weeks — The agent needs baseline eating data to make meaningful recommendations. Sporadic logging produces unreliable adjustments. Two consistent weeks establishes your actual intake patterns.

5. Report injuries and discomfort immediately — Working through pain creates worse injuries. The agent can modify exercises around any limitation, but only if it knows about the issue. Early reporting leads to better exercise substitutions.

Common Issues

Wearable data seems inaccurate or missing

Ensure your device firmware is updated and sync is enabled for the specific metrics the agent reads (HRV, sleep, resting HR). Apple Health requires explicit permission for each data type in Settings > Health > Data Access. If data gaps persist, the agent falls back to RPE-based recovery assessment.

Calorie targets feel too high or too low

The initial calculation uses standard metabolic formulas which have a margin of error. Track your weight for 2 weeks at the suggested intake. If weight changes faster or slower than predicted, tell the agent and it will recalibrate your TDEE estimate based on actual results rather than formulas.

Program does not match my gym's equipment

Specify exact equipment in the equipment array rather than using "full_gym." If your gym lacks specific machines, list what you have: ["barbell", "dumbbells", "cable_machine", "pull_up_bar"]. The agent will substitute exercises that require unavailable equipment.

Privacy & Data Handling

All fitness data, training logs, and nutrition information are stored locally on your device. Wearable integrations use read-only API access to import health metrics — the agent never writes data back to Apple Health, Fitbit, or Garmin. No body measurements, health metrics, or dietary information are transmitted to external servers. MyFitnessPal integration uses their public API with your explicit authorization and only reads meal log summaries. You retain full control over all health data and can revoke platform access at any time through each service's settings. Generated workout plans and nutrition targets are saved as local files that you can delete at will.