About this project

FitBuddy is an open-source Android app for tracking fitness and nutrition with AI assistance. Users can log meals and workouts using a photo or loose text, and a configured LLM estimates calories and macronutrients. The app includes a daily dashboard with a calorie ring, food and exercise logs, macro breakdowns, progress charts, editable meal reviews with live macro recalculation, reusable food presets, and science-based body goal tracking for weight, body fat, BMR, and muscle mass. Region packs let users personalize AI prompts, staple foods, portion language, and log examples. AI providers are configurable at runtime: OpenRouter, Gemini, OpenAI, and Ollama or other OpenAI-compatible endpoints. Without a key, an offline simulator provides fallback functionality. Data is stored locally, and backups can be encrypted with AES-256-GCM and exported locally or to the cloud. The app supports Material You dynamic theming on Android 12+, optional anonymous crash reporting via Sentry, and F-Droid distribution. The project is built with Kotlin, Jetpack Compose, Room, Retrofit, and DataStore using an MVVM architecture. CI builds and tests on every push, and release workflows produce signed APKs and AABs. The source is licensed under GPL-3.0.