Diet Mate

A personal nutrition tracking web app that syncs with Fatsecret and helps maintain healthy eating habits without a coach.

What it does

• Pulls daily food diary from Fatsecret via OAuth 1.0
• Categorizes food items using GPT-4o-mini with local caching
• Tracks 7 nutrition parameters with progress bars
• Manual daily input for weight and steps
• Weekly report showing trends and goal hit rate

Tracked parameters

Parameter	Unit
Calories	kcal/day
Vegetables, fruits & greens	g/day
Avocado	g/day
Calcium	mg/day
Omega-3	g/day
Eggs	pcs/day
Seafood	portions/day

All targets are configurable in the Settings screen.

Tech stack

• Next.js — frontend + API routes
• Fatsecret API — food diary (OAuth 1.0)
• OpenAI GPT-4o-mini — food categorization
• USDA FoodData Central — Omega-3 data
• SQLite (better-sqlite3) — local storage

Setup

1. Clone the repo and install dependencies:

   npm install

2. Create .env.local:

   FATSECRET_CLIENT_ID=your_client_id
   FATSECRET_CLIENT_SECRET=your_oauth2_client_secret
   FATSECRET_CONSUMER_SECRET=your_oauth1_consumer_secret
   OPENAI_API_KEY=your_openai_key
   USDA_API_KEY=your_usda_key        # optional, DEMO_KEY used if empty
   NEXTAUTH_URL=http://localhost:3000
   NEXTAUTH_SECRET=your_random_secret
   

3. Run the dev server:

   npm run dev

4. Authorize Fatsecret by opening:

   http://localhost:3000/api/auth/start
   

5. Open the app at http://localhost:3000

Notes

• Fatsecret food diary access requires OAuth 1.0 via authentication.fatsecret.com
• Food categories are cached in SQLite after the first classification — no repeated API calls
• The database file diet-mate.db is created automatically in the project root