FitMatch AI

App System authentication integration and rollout instructions: docs/app-system-auth.md.

FitMatch AI is a production-ready Next.js App Router MVP for AI-powered fashion recommendations. Users can:

• sign in with Microsoft via Supabase Auth
• upload or capture clothing photos
• extract item metadata with OpenAI vision
• save items into a private digital wardrobe
• take a style quiz and store a persistent profile
• generate multi-style outfit suggestions
• save generated outfits for later review

Stack

• Next.js 16 App Router + TypeScript
• Tailwind CSS 4
• Supabase Auth, Database, and Storage
• OpenAI API for image analysis and outfit generation

Environment variables

Copy .env.local.example to .env.local and fill in:

NEXT_PUBLIC_SUPABASE_URL=
NEXT_PUBLIC_SUPABASE_ANON_KEY=
SUPABASE_SERVICE_ROLE_KEY=
OPENAI_API_KEY=
OPENAI_TEXT_MODEL=gpt-4o-mini
OPENAI_VISION_MODEL=gpt-4o-mini
OPENAI_IMAGE_MODEL=gpt-image-1

Supabase setup

1. Run the SQL in supabase/schema.sql in your Supabase SQL editor.
2. In Supabase Auth, enable the Microsoft provider. Supabase labels this provider as azure.
3. Add the callback URL:

https://YOUR_DOMAIN/auth/callback
http://localhost:3000/auth/callback

4. Confirm the private storage bucket fitmatch-wardrobe exists if you manage buckets manually.
5. Add your Supabase service role key to SUPABASE_SERVICE_ROLE_KEY so server-side uploads can write to Storage reliably.

Local development

npm install
npm run dev

Then open http://localhost:3000.

Production checks

The app currently passes:

npm run lint
npm run build

API routes

• POST /api/analyze-image
• POST /api/generate-outfit
• POST /api/generate-outfit-image
• GET/POST /api/user/profile
• GET/POST /api/wardrobe
• PATCH/DELETE /api/wardrobe/[id]
• GET/POST /api/outfits

Notes

• OpenAI API errors (missing key, quota exceeded) are logged and surfaced transparently so developers and users receive clear actionable feedback.
• Wardrobe and AI lookbook images are securely stored in private Supabase Storage buckets and served via signed URLs.