LifeMog

Personal health tracker for meals, workouts, and body weight. Mobile-first PWA with AI-powered food logging.

Features

Food — Log meals by describing them in natural language or uploading photos. Tracks calories, protein, carbs, and fat against daily goals.

Workout — Log exercises with sets, reps, and weight. Visual muscle map shows which muscles you've trained, color-coded by volume. Navigate day-by-day with swipe or arrows.

Calendar — Browse past days to review meals and workouts.

Body Weight — Log daily weigh-ins with unit toggle (kg/lbs). History graph and entry management.

AI Meal Analysis — Describe a meal or take a photo and get instant macro estimates. Uses Gemini 2.5 Flash (vision) by default, with MiniMax as a fallback.

Stack

• Vanilla HTML/CSS/JS — no framework, no build step
• Firebase Auth (Google Sign-In) + Firestore (per-user data)
• Netlify Functions — serverless AI proxy
• Gemini 2.5 Flash (primary) / MiniMax Text-01 (fallback)

Setup

1. Clone

git clone https://github.com/ish-cs/lifemog.git
cd lifemog

2. Firebase

1. Create a Firebase project at console.firebase.google.com
2. Enable Authentication → Google sign-in
3. Enable Firestore (start in production mode)
4. Copy your web app config into index.html at the firebase.initializeApp({...}) call

3. Environment variables

Create a .env file (for local Netlify dev) or set in Netlify dashboard:

GEMINI_API_KEY=your_gemini_api_key
FIREBASE_API_KEY=your_firebase_api_key

FIREBASE_API_KEY is used server-side to verify ID tokens. GEMINI_API_KEY powers AI meal analysis.

MiniMax is optional — users can enter their own key in Settings.

4. Deploy

npm install -g netlify-cli
netlify dev       # local dev
netlify deploy    # preview
netlify deploy --prod

Firestore Rules

All user data is scoped to users/{uid} — no user can read or write another's data.

match /users/{uid}/{document=**} {
  allow read, write: if request.auth != null && request.auth.uid == uid;
}