# V97 RC10 – Batch 04 Greenscreen Import QA

Batch 04 was generated in-chat using `next-athlete-greenscreen-16x9-v4`, then processed through the reusable greenscreen keying pipeline.

## Imported exercises
- `carioca-shuffle` → `assets/img/exercises/figures/carioca-shuffle.webp`
- `med-ball-slam` → `assets/img/exercises/figures/med-ball-slam.webp`
- `goblet-squat` → `assets/img/exercises/figures/goblet-squat.webp`
- `chin-up` → `assets/img/exercises/figures/chin-up.webp`
- `db-farmer-carry` → `assets/img/exercises/figures/db-farmer-carry.webp`
- `plank` → `assets/img/exercises/figures/plank.webp`

## Pipeline
1. Raw greenscreen sources copied to `incoming/greenscreen-source/`.
2. Batch 04 was keyed to alpha WebP with the greenscreen keying helper.
3. QA composites were created in `qa_screenshots/` for black, blue and light backgrounds.
4. `tools/import-approved-figures.py` copied approved WebPs into `assets/img/exercises/figures/` and updated `status.json`.

## Validation
- All six Batch 04 outputs are RGBA WebP.
- All six Batch 04 outputs are 1600×900.
- Alpha channel range verified as 0–255.
- Smoke tests run: RC122, RC123, RC124.

## Remaining / tomorrow
- Browser/app visual QA was not run for this package; static alpha/composite QA files are included.
- Batch 05 is intentionally left as `batch-05-next` with empty items so the next six exercises can be selected tomorrow.