A precise facial aesthetics analyzer powered by Google's Gemini 2.0 Flash AI. Upload a photo or use your camera to receive an in-depth PSL (Pretty Scale of Looks) diagnostic report including feature-by-feature breakdowns, written analysis, and actionable looksmaxxing tips.

• Real-time Camera Capture — Use your device camera to take photos directly in the app
• Image Upload — Upload photos from your device or clipboard
• PSL Score Analysis — Get an accurate PSL score with percentile and tier classification
• Feature Ratings — Detailed 1–10 scores for 9 facial features:
  • Jawline
  • Cheekbones
  • Canthal tilt
  • Eye area
  • Nose
  • Chin
  • Symmetry
  • Facial thirds
  • Skin quality
• Written Analysis — Comprehensive breakdown of facial structure and harmony
• Strengths & Weaknesses — Bulleted lists identifying key assets and areas for improvement
• Looksmaxxing Tips — Actionable suggestions for maximizing facial aesthetics through methods like skincare, haircuts, styling, and fitness

• Frontend Framework — React 19 with TypeScript
• Build Tool — Vite 6
• Styling — Tailwind CSS 4 with Vite plugin
• AI Model — Google Gemini 2.0 Flash via @google/generative-ai
• Linting — ESLint 9 with TypeScript support

• Node.js 18+ and npm
• A Google Gemini API key (get one free)

# Install dependencies
npm install

# Start the development server
npm run dev

The app will open at http://localhost:5173 (default Vite port).

npm run build
npm run preview  # Preview the production build locally

1. Paste Your API Key — Enter your Google Gemini API key (stored only in your browser's local storage; never sent to servers)
2. Add a Photo — Either:
   • Click "Upload image" to select a file
   • Click "Use camera" to capture with your device camera
3. Run Analysis — Click "Run analysis" to submit
4. Review Results — Get your PSL score, feature ratings, analysis, and tips

src/
├── App.tsx                 # Main app component
├── index.css               # Tailwind CSS configuration
├── main.tsx                # React root
├── components/
│   └── LoadingSpinner.tsx   # Loading indicator
├── lib/
│   ├── gemini.ts           # Gemini API integration
│   └── scoreColors.ts      # Color coding logic for scores
└── types/
    └── psl.ts              # TypeScript types and system prompt

The app runs entirely in the browser. Your API key is:

• Stored in browser local storage (not uploaded anywhere)
• Used only to call Google's Gemini API directly from your device
• Never logged or persisted by this application

npm run lint  # Check for code issues

MIT

Built with React, Vite, Tailwind CSS, and Google Gemini AI for Sillyhacks 2026.