Good morning πŸ‘‹

Let's build the habit today.

πŸ”₯ streak 0 days
Workouts logged
0
all time
Volume lifted
0kg
sets Γ— reps Γ— weight
Today's calories
0kcal
Protein today
0g

🎯 Today's plan

⚑ Quick log

πŸ•˜ Recent activity

Exercise Library

Pick a muscle β†’ get animated demonstrations, form cues and step-by-step technique.

β€”

πŸ«€ Target a muscle group

My Journey

Every session, every kilo, every PR β€” tracked on your device.

Sessions
0
Total volume
0kg
Total minutes
0
Best streak
0d

πŸ“… Last 14 days

πŸ† Personal records

πŸ“‹ Session log

Indian Nutrition

Real desi food, real macros. Build a plate that fits your goal.

Goal: lean

πŸ₯— Today's Indian meal plan

πŸ› Indian food database per standard serving

πŸ’‘ Coach's tips for Indian eating

Ask AI Coach

Food suggestions, workout swaps, plateau fixes β€” grounded in Indian food & your own logs.

model built-in
FitMitra Coach

πŸ”Œ Use a real LLM (optional)

Works with Ollama (ollama serve), LM Studio, llama.cpp server, or any OpenAI-style API. Leave blank to keep using the offline coach β€” no data leaves your device.

**FitMitra** is live in the side panel β€” a complete single-file webapp, no build step, no backend. **How to use it locally:** save the file as `fitmitra.html` and open it in any browser (or run `python3 -m http.server` in the folder and visit `localhost:8000/fitmitra.html`). ### What's in it **πŸ‹οΈ Exercise Library (10 muscle groups, 39 exercises)** Chest, back, shoulders, biceps, triceps, legs, glutes, core, cardio, full-body. Each exercise card shows **two demonstration stills pulled live from the open wger exercise database** (CC-BY-SA 4.0, no API key), a hover play overlay, and a detail modal with numbered step-by-step technique, a coach's cue, equipment/level/suggested sets-reps, a "next exercise" button, and a YouTube form-tutorial link. If offline or blocked, a generated SVG muscle card renders instead β€” nothing ever breaks. **πŸ“ˆ Journey tracking** Quick-log (exercise, minutes, sets, reps, kg), per-exercise **PR table**, a 14-day training heatmap, current + best streak, total volume lifted, full session log with delete, and JSON export/reset. **πŸ› Indian Nutrition** 70-item Indian food database with realistic per-serving **kcal / protein / carbs / fat** β€” roti, buttery naan, poha, idli, masala dosa, rajma, chole, dal tadka, palak paneer, soya chunks, besan chilla, curd, lassi, chicken biryani, samosa, gulab jamun. A meal-plan generator builds breakfast/lunch/pre-workout/dinner around your goal (**fat loss, lean gain, pure veg, high protein, budget/hostel**) with live macro totals and a one-tap "Eat this" logger. Plus 8 desi coaching tips (the plate rule, the 20-minute rule, chai calories, protein per rupee). **πŸ€– AI Coach** Works instantly with a **built-in offline coach** β€” it parses your question for goal/diet/budget keywords and answers by ranking the Indian food database against your *actual logged macros*, so "I'm vegetarian, need 100g protein under 1800 kcal" produces real ranked dishes with numbers. Optionally connect **Ollama, LM Studio, llama.cpp or any OpenAI-compatible endpoint** in the box at the bottom β€” it then injects your streak, recent sessions and today's macros as context. If the endpoint fails, it silently falls back offline. **Two things to know:** - Exercise images are fetched from `wger.de` at runtime, so the library looks best with internet; the SVG fallback covers offline use. - Since the sandbox has no `localStorage`, data lives in memory for the session β€” use **Export JSON** in My Journey to keep your history. Want me to add a set-by-set live workout mode with rest timers, a body-weight/measurement tracker with a progress chart, or swap the wger images for true embedded animated GIFs?