Fire TV · Vega OS
Everybody Moves
One family workout on the TV, where each person gets their own version of every move: a chair squat for Grandpa, a glute bridge for sore knees, jump squats for whoever is ready, all at the same time.
Joining a session? Type the four-letter room code shown on the TV.
How it works
- The TV shows a QR code. Everyone checks in from their phone: energy from 1 to 5, and anything sore (knees, back, shoulders, wrists). People without a phone join with the remote.
- The coach plans today's session from those check-ins and the household's history: the focus is whatever was trained least recently, and the length follows the family's pattern at this time of day.
- During the workout, the TV shows each person's version of the move, with a reason ("easy on the knees"). Phones show your own version, the same timer, and captions of the coach.
- Afterwards, everyone rates it. "Easy" twice in a row moves you up a level; "hard" moves you down.
Try the TV app
The app runs on Fire TV's Vega OS. With the Vega SDK installed:
git clone https://github.com/danielhagever/everybody-moves cd everybody-moves/tv npm install npm run build:debug vega virtual-device start vega run-app build/aarch64-debug/everybodymoves_aarch64.vpkg com.glitchbound.everybodymoves.main -d VirtualDevice
It opens on a sample household with two weeks of history. On the "Who's working out?" screen, set Pace: demo (5x) to see a whole session in about two minutes, and open the room's link on your phone to check in.
Built with
- React Native for Vega, the Vega W3C media AudioPlayer (the coach's voice), react-native-svg (figures, countdown ring, QR code)
- Cloudflare Workers and D1 (planner, sessions, phone check-in), Workers AI: Llama 4 Scout (the coach's introduction, checked by rules before it's spoken) and Deepgram Aura-2 (voice)
Code: github.com/danielhagever/everybody-moves · General fitness guidance, not medical advice.