Shipped
- Engine behavior overhaul (stability + correctness pass)
- Load kind classification (barbell/dumbbell/etc.)
- Removed dropset concept from UX to align with schema expectations
- Build/validation hardening fixes
Why
- Schema drift is silent corruption and kills long-term maintainability
- Load-kind metadata improves programming precision and future analytics
Notes
- This release prioritizes correctness over feature theatre.
Next
- Onboarding and identity layer
- Quick workouts for a low-friction “just train” entry point