CodeWithCJ/SparkyFitness

SparkyFitness is a self-hosted fitness and health tracking platform for households that want nutrition, exercise, hydration, sleep, fasting, mood and body measurements in one place instead of across several subscriptions. It ships a backend server, web frontend, iOS/Android apps and Apple Watch/Wear OS companions, and keeps the stored data on infrastructure you control SparkyFitness repository.

What you actually get

You log food, workouts, water, weight and other daily check-ins, set goals, and review charts and custom date-range reports across long histories. Multiple profiles with granular family-sharing permissions are supported, along with OIDC, TOTP, passkey and MFA sign-in options. Sleep stages, HRV and blood glucose are not measured by the app; they are imported from a watch, phone or meter and then stored and charted project README.

How data gets in

The platform connects to a long list of sources: Apple Health and Google Health Connect on phones, plus Fitbit, Garmin Connect, Withings, Polar Flow, Oura, Strava, Hevy and several food databases such as OpenFoodFacts, USDA and Mealie. Once connected, steps, workouts, sleep and body measurements sync automatically to your server project README. Some integrations are explicitly marked as partially tested (Strava) or reliant on an unofficial API (Yazio), so treat those as less predictable than the mainstream ones.

Optional AI and the install choice

SparkyAI, a chatbot for logging food, exercise and body stats by chat and for uploading meal photos, is in beta. There are two documented ways to run the service: a Docker Compose self-hosted install, or a managed deployment through third-party providers PikaPods or Zenith for people who don't want to administer a server project README. The self-hosted path requires you to supply your own database password, API encryption key, auth secret and public frontend URL; the server refuses to start while the published placeholder values remain. Full configuration for SSO, SMTP, external databases and reverse proxies is in the documentation site installation guide.

Constraints worth knowing

The project describes itself as source-available, not open source: the licence is non-commercial and commercial use needs permission license. It also states it is under active development, that breaking changes may occur between releases, and that auto-updating containers are not recommended; review release notes before upgrading. The AI chatbot, family/friends access and API documentation are all labelled beta project README.