Skip to content

CLAUDE.md

Latest commit

 

History

History
52 lines (34 loc) · 2.64 KB

CLAUDE.md

File metadata and controls

52 lines (34 loc) · 2.64 KB
Raw

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

Appetight is a hackathon iOS nutrition tracking app that uses Claude AI for food image analysis, restaurant recommendations, and gym busy-time predictions. Users log meals via camera (Claude analyzes food photos), voice, or manual input.

Build & Run

This is a native iOS app — all development happens in Xcode.

  • Build: Open Appetight.xcodeproj → ⌘B
  • Run: ⌘R (requires physical device or simulator)
  • Test: ⌘U (no tests currently exist)

Before first build, add API keys to Appetight/Config.swift:

static let anthropic = "sk-ant-..."
static let googleMaps = "AIzaSy..."

Architecture

Central state: AppState.swift is an @EnvironmentObject injected at root. It holds the user profile, all meal logs, friends list, and streak data, and persists everything to UserDefaults.

Claude integration: AnthropicService.swift makes all Anthropic API calls. Uses claude-haiku-4-5-20251001. Three main call sites:

  1. Food image analysis (base64 image → nutrition facts)
  2. Restaurant recommendations (location + persona context → ranked list)
  3. Gym busy-time prediction (gym name + hour → estimated crowding)

Personalization: PersonaEngine.swift aggregates meal logs into a UserPersona (SwiftData model in UserPersona.swift). The persona is serialized into a string and injected into Claude system prompts to personalize responses over time.

External data: PlacesService.swift calls Google Maps Places API for nearby restaurants and gyms. LocationService.swift wraps CoreLocation.

Navigation flow:

  • AppetightApp.swift → RootView checks for saved profile → routes to OnboardingView or DashboardView
  • DashboardView is a 4-tab container: Today, Hungry, Gym, Friends
  • Meal logging opens modally from the Today tab (camera, voice, or quick-add)

Data models: All core types (Profile, MealLog, Restaurant, Friend, etc.) are defined in Models.swift. TDEE/macro calculations live in TDEE.swift (Mifflin-St Jeor formula).

Key Patterns

  • MVVM with @StateObject / @EnvironmentObject / @Observable
  • async/await throughout all service calls
  • No external Swift packages — only Apple frameworks (SwiftUI, SwiftData, CoreLocation, Speech, AVFoundation, UserNotifications)
  • SwiftData used only for UserPersona; everything else persists via UserDefaults

Required Permissions

The app requests camera, microphone, location, speech recognition, photo library, and notifications at runtime. All permission strings must be present in Info.plist.