Woof

Woof is an Expo/React Native mobile app for scanning pet food and turning product labels, barcodes, and food photos into structured nutrition analysis. It combines mobile camera flows, Supabase-backed auth/history, and server-side AI analysis behind an Edge Function.

Product Scope

• Scan pet food by barcode or photo.
• Analyze verified product data when available through Open Pet Food Facts.
• Fall back to AI-assisted analysis for product photos and human-food safety questions.
• Stream partial results into the UI so long-running analysis does not feel frozen.
• Persist scan history locally and sync it to Supabase for signed-in users.
• Gate premium flows through RevenueCat subscription integration.

Tech Stack

• Expo and React Native
• React Navigation
• Supabase Auth, Postgres, Row Level Security, and Edge Functions
• Claude API through a server-side Supabase Edge Function
• Open Pet Food Facts lookup
• AsyncStorage for local persistence
• RevenueCat for subscriptions

Architecture

App.js
|-- screens/                  # Auth, onboarding, scanner, results, profile, paywall
|-- services/
|   |-- analysisService.js    # Background analysis singleton and pub/sub state
|   |-- claude.js             # Streaming API client
|   |-- opff.js               # Open Pet Food Facts lookup
|   |-- cache.js              # Supabase-backed analysis cache
|   |-- history.js            # Local and remote scan history
|   `-- purchases.js          # RevenueCat integration
`-- supabase/
    |-- functions/analyze/    # Server-side AI proxy and scoring prompts
    `-- migrations/           # Auth, history, rate limit, cache, and cleanup schema

Notable Implementation Details

• Background analysis survives screen unmounts and supports multiple UI subscribers.
• Barcode results check cache and verified product data before invoking AI.
• Photo analysis uses a temporary key that is re-keyed once the product identity is known.
• Supabase RLS policies keep profile and history data scoped to the current user.
• The Edge Function keeps AI provider secrets off the mobile client.

Local Setup

npm install
npm start

Create a local .env based on .env.example:

SUPABASE_URL=
SUPABASE_ANON_KEY=
GOOGLE_WEB_CLIENT_ID=
REVENUECAT_API_KEY_IOS=
REVENUECAT_API_KEY_ANDROID=

Deploy the Supabase Edge Function and migrations before testing full analysis flows.

Quality Notes

The repository includes an audit document covering architecture, security, accessibility, and operational follow-ups. The public .env.example file is safe to commit; real keys should stay in local environment files or hosted secret managers.

Quality Signals

• CI installs dependencies and verifies the expected app, Supabase, docs, and migration structure.
• CONTRIBUTING.md documents the expected validation path for scanner, analysis, auth, and migration changes.
• SECURITY.md captures credential, Edge Function, RLS, and privacy expectations.