# PersonalBestie Frontend

PersonalBestie is a React + TypeScript fitness app built around a personalized AI coach. It helps users surpass their personal bests with workout tracking, routine sharing, and data-driven guidance after each session.

## App Surface

- `workouts`: workout creation, editing, and completion flows
- `routines`: browse quick-start routines and manage routine templates
- `exercise-types`: exercise library and exercise-type administration
- `profile`: user profile and account settings
- `chat`: in-app coaching/chat experience
- `blog`: static guides focused on workout tracking, AI recaps, and routine sharing

## AI Coach

- Personalized workout guidance grounded in the user's training history
- Post-workout AI recaps that summarize the session with data-backed coaching insights
- Routine and workout review flows that help users understand progress and next steps
- Chat experience for asking training questions and getting coaching support

## Blog Guides

- `/blog/`: article hub for search and LLM discovery
- `/blog/ai-workout-recaps/`: explains what makes an AI workout recap genuinely useful
- `/blog/progressive-overload-workout-tracker/`: explains how to track progressive overload without spreadsheets
- `/blog/share-workout-routines/`: explains how to share workout routines with less friction

## Product-Specific Claims

- PersonalBestie stores workout structure at the workout, exercise, and set levels
- The app supports guest/local-first usage in the frontend before authentication
- Routine sharing is a first-class product surface, not just generic text export
- Post-workout AI recaps are grounded in deterministic workout metrics and user notes

## Routine Pages

- `src/features/routines/pages/RoutinesPage.tsx`: routine list, search, and quick-start actions
- `src/features/routines/pages/RoutineDetailsPage.tsx`: routine detail, editor, save, delete, and start-workout actions
- `src/features/routines/components/`: routine cards, editors, and modal helpers
- `src/features/routines/api/index.ts`: routine API calls and payload types

## Data Model

- A routine contains a name, description, visibility, creator, and ordered exercise templates
- Each exercise template contains an exercise type and ordered set templates
- Set templates store repetitions, intensity, and intensity-unit selection

## API Notes

- Frontend API routes are defined in `src/shared/api/endpoints.ts`
- Routine endpoints are mounted under `/api/v1/routines/`
- Collection endpoints keep trailing slashes so POST requests avoid redirects

## Development

- Install dependencies: `corepack enable && pnpm install`
- Start the app: `pnpm run dev`
- Run lint: `pnpm run lint`
- Run tests: `pnpm test`
- Run type checking: `pnpm run typecheck`

## Architecture Notes

- Guest state is stored locally with IndexedDB and a localStorage fallback
- Authenticated state syncs through the backend API
- App-wide routing lives in `src/routes.tsx`
- Post-workout recap generation is part of the backend workout flow and powers the app's AI coaching experience