SnapMacros

Honest AI nutrition tracking: Track food in 5 seconds. Private by default. No guilt, no fake precision.

SnapMacros is a mobile app MVP for fast meal photo logging with confidence ranges, one-question correction flows, reusable common meals, and a low-cost subscription wedge.

MVP status

Scaffolded Expo app with:

• Daily calories/macros dashboard
• Photo capture or library upload flow
• Optional meal description / voice-ready text correction field
• Local nutrition estimate engine with confidence ranges
• One-tap oil/butter/sauce correction
• Local meal history via AsyncStorage
• Product settings screen with pricing and privacy positioning

The first build uses a deterministic local estimator so the app can be tested immediately. The AI layer is isolated behind app/lib/estimate.ts so it can be replaced with Gemma 3n / cloud fallback later without rewriting the UI.

Run locally

cd Projects/SnapMacros/app
bun install
bun start

Then scan the Expo QR code with your phone, or run the iOS/Android simulator from Expo.

Product direction

• Free: 10 scans/week
• Paid: $29–39/year
• Family: $59/year
• Optional one-time packs: dietitian audit, 7-day meal plan

Project structure

Projects/SnapMacros/
├── README.md
├── PRODUCT_SPEC.md
├── docs/
│   └── 2026-05-18-snapmacros-mvp-design.md
└── app/
    ├── app/                 Expo Router screens
    ├── components/          Shared UI
    ├── lib/                 Estimate engine, storage, types, theme
    ├── assets/images/       App icon/splash
    ├── app.json
    ├── package.json
    └── tsconfig.json