The FANG Framework: How to Make AI Your Fitness Coach 🧠
Generic fitness advice is why 95% of programs fail. "Do 3 sets of 10" means nothing without context — YOUR training age, YOUR recovery capacity, YOUR equipment, YOUR goals.
The FANG Framework gives AI everything it needs to build programs that work specifically for you:
- Frequency — How often you train and what your schedule allows
- Adaptation — Your current fitness level, training history, and how your body responds
- Nutrition — Your dietary context, caloric needs, and food constraints
- Goals — What you're actually optimizing for, prioritized and time-bound
Every prompt in this guide uses FANG. Master it once, use it forever.
Module 1: Know Your Training Profile 🏋️
Before AI can help you, it needs to understand WHERE you are. Not where you think you are — where you actually are.
The AI Training Assessment
Use this prompt to establish your baseline:
I need a comprehensive training assessment. Here's my profile:
Age: [age], Sex: [M/F], Height: [height], Weight: [weight]
Body fat estimate: [if known, or "unsure"]
Training history: [how long you've been training consistently]
Current program: [what you're doing now, or "nothing consistent"]
Best lifts (if applicable): Bench [X], Squat [X], Deadlift [X], OHP [X]
Cardio baseline: [can run X miles, or "minimal cardio fitness"]
Injuries/limitations: [list everything — old and current]
Sleep: [average hours, quality 1-10]
Stress level: [1-10]
Available equipment: [home dumbbells / full gym / etc.]
Training days available: [X days per week, X minutes per session]
Based on this, classify my training level (beginner/intermediate/advanced), estimate my recovery capacity, and identify my biggest limiting factors for progress.Training Level Classification
AI should classify you honestly — not flatteringly. Here's the reality check:
| Level | Timeframe | Strength Indicators (Male, ~180lb) | What It Means |
|---|---|---|---|
| Beginner | 0-12 months consistent | Bench <135, Squat <185, DL <225 | Linear progression works. Add weight every session. |
| Intermediate | 1-3 years | Bench 185-275, Squat 275-405, DL 315-455 | Need periodization. Can't add weight every session. |
| Advanced | 3-5+ years | Bench >275, Squat >405, DL >455 | Need complex programming. Gains are slow and hard-won. |
Module 2: Workout Programming 📋
This is where AI shines. A good prompt produces programming that would cost $200+ from a certified strength coach.
The Program Design Prompt
Design a complete training program using this FANG profile:
FREQUENCY: [X] days/week, [X] minutes/session, [specific days available]
ADAPTATION: [training level], [years training], [current lifts/fitness level], [injury history], [recovery: sleep X hrs, stress X/10]
NUTRITION: [current caloric intake if known], [diet type], [willing to track macros: yes/no]
GOALS: Primary: [goal 1]. Secondary: [goal 2]. Timeline: [X weeks/months].
Requirements:
- Include warm-up protocol (5 min)
- Main compound movements with sets, reps, RPE/RIR targets
- Accessory work with specific rep ranges
- Progressive overload scheme (how to add weight/volume over time)
- Deload protocol (every X weeks)
- Cardio prescription if applicable
- Total program length with phasesProgramming Concepts AI Handles Well
Progressive Overload — AI can model multiple overload strategies:
- Weight progression (add 5lb upper, 10lb lower each cycle)
- Volume progression (add 1 set per movement per week)
- Density progression (same work in less time)
- RPE progression (same weight, lower perceived effort = strength gain)
Periodization — Ask AI to structure phases:
Structure this 16-week program in 4 phases:
Phase 1 (weeks 1-4): Hypertrophy — higher volume, moderate intensity (65-75% 1RM)
Phase 2 (weeks 5-8): Strength — moderate volume, higher intensity (75-85% 1RM)
Phase 3 (weeks 9-12): Peak — lower volume, high intensity (85-95% 1RM)
Phase 4 (weeks 13-16): Deload + test new maxes, then repeat
Include a deload week at the end of each phase.Exercise Selection — AI knows hundreds of exercises and their muscle activation patterns. But always validate:
For each exercise in this program, tell me: (1) primary muscles worked, (2) common form mistakes, (3) a regression if I can't do it, and (4) a progression when it gets easy.Module 3: Nutrition Architecture 🥗
Training is the stimulus. Nutrition is the result. AI can build meal plans that most dietitians would charge $200+ for — IF you give it the right context.
Step 1: Calculate Your Targets
Calculate my daily caloric and macro targets:
Stats: [age], [sex], [height], [weight], [body fat % if known]
Activity level: [sedentary job + X training sessions/week, or describe daily activity]
Goal: [fat loss / maintenance / lean bulk / aggressive bulk / recomp]
Rate of change: [lose X lb/week OR gain X lb/week]
Use the Mifflin-St Jeor equation for BMR, then apply appropriate activity multiplier. Give me:
1. Daily calorie target
2. Protein target (g and g/lb bodyweight)
3. Fat target (g and % of calories)
4. Carb target (remaining calories)
5. Fiber target
6. Water intake recommendationStep 2: Build the Meal Plan
Build a 7-day meal plan matching these targets: [calories]cal, [P]g protein, [C]g carbs, [F]g fat.
Constraints:
- Dietary restrictions: [vegan / vegetarian / gluten-free / none / allergies]
- Budget: $[X]/week on groceries
- Cooking skill: [beginner / intermediate / advanced]
- Max prep time per meal: [X] minutes
- Meals per day: [X] (include snacks if applicable)
- Meal prep: [yes — batch cook Sundays / no — cook daily]
- Foods I dislike: [list]
- Foods I love: [list]
For each day, provide: meal name, ingredients with gram weights, macros per meal, and running daily totals. End with a consolidated grocery list organized by store section.The Nutrition Evidence Hierarchy
AI sometimes conflates trendy nutrition advice with evidence-based nutrition. Use this framework:
| Evidence Level | What It Means | Examples |
|---|---|---|
| 🟢 Strong (meta-analyses, RCTs) | Trust this | Caloric deficit = fat loss, protein 0.7-1g/lb for muscle, creatine works |
| 🟡 Moderate (multiple studies, some conflicting) | Likely true, but context matters | Meal timing around workouts, carb cycling for body comp, sleep affecting fat loss vs muscle loss |
| 🟠 Emerging (limited studies, mechanistic reasoning) | Interesting but don't bet on it | Specific microbiome interventions, precise meal frequency effects, cold exposure for fat loss |
| 🔴 Weak/None (anecdotal, influencer claims) | Ignore unless proven otherwise | "Eating clean" without caloric context, detox diets, most fat burners, alkaline diet |
Prompt: Evaluate these nutrition claims using the evidence hierarchy (strong/moderate/emerging/weak). For each, cite the level of evidence and one key study or meta-analysis: [list your claims or questions]Module 4: Recovery & Adaptation 😴
Training breaks you down. Recovery builds you back stronger. AI can optimize recovery — but most people never think to ask.
Sleep Optimization
I sleep [X] hours per night. My sleep quality is [X/10]. I [do/don't] wake up during the night. My biggest sleep issues are: [falling asleep / staying asleep / waking too early / poor quality despite hours].
What specific, actionable changes should I make to improve sleep for athletic recovery? Include: timing recommendations, environment changes, supplement options with evidence ratings, and pre-sleep routine.Recovery Protocol Design
Design a recovery protocol for my training schedule:
- Training days: [list days and types — e.g., Mon: heavy squat, Tue: upper hypertrophy...]
- Current recovery methods: [stretching / foam rolling / sauna / cold plunge / nothing]
- Pain points: [sore areas, tight areas, areas that feel "off"]
- Time available for recovery: [X minutes per day]
Include: daily mobility routine (5-10 min), post-workout recovery protocol, active recovery day programming, and when to take a full rest day vs. active recovery.Deload Strategies
AI should know when to pull back. If you're not asking about deloads, you're probably overtraining.
I've been training [X weeks] without a deload. My performance has [improved / plateaued / declined] over the last [X] sessions. My motivation is [high / low / inconsistent]. My sleep is [good / disrupted].
Do I need a deload? If yes: prescribe a deload week — should I reduce volume, intensity, or both? By how much? Should I maintain frequency?Module 5: Special Populations & Goals 🎯
Returning from Injury
I'm returning from [injury type] that occurred [timeframe] ago. My doctor/physio has cleared me for [activities cleared]. Pain level is [0-10]. What I can't do yet: [movements to avoid].
Design a 4-8 week return-to-training program that progressively reintroduces [movements] while protecting [injured area]. Include mobility/rehab work for [area] in every session.Body Recomposition (Lose Fat + Build Muscle)
I want to recomp (lose fat and build muscle simultaneously). Stats: [weight], [body fat %], [training level]. Is recomp realistic for me? If yes, design the nutrition and training approach. If no, should I cut or bulk first, and why?Recomp reality check — AI should tell you the truth:
- Beginners (<1 year training): Recomp works very well
- Intermediate (1-3 years): Recomp is slow, cut/bulk cycles are faster
- Advanced (3+ years): Recomp is almost impossible, cut or bulk
- Overfat beginners: Recomp is the ideal strategy — don't cut, just train and eat at maintenance with high protein
Endurance Training
Build a [5K / 10K / half marathon / marathon] training plan:
Current fitness: [can run X miles at X pace, or "starting from zero"]
Goal race: [distance] in [time goal, or "just finish"]
Available running days: [X per week]
Cross-training available: [days/types]
Injury history: [especially knees, ankles, hips, IT band]
Timeline to race: [X weeks]
Include: weekly mileage plan with easy/tempo/interval breakdown, cross-training schedule, taper protocol, and fueling strategy for race day.The FANG Checklist
Before you send ANY fitness prompt, verify:
- [ ] Frequency — Did I specify training days, session length, and schedule?
- [ ] Adaptation — Did I include training level, current numbers, injuries, and recovery?
- [ ] Nutrition — Did I include caloric context, diet type, and food constraints?
- [ ] Goals — Did I prioritize goals and set a timeline?
Miss any of these and AI will fill in assumptions — which may be wrong for YOUR body.