Flower Factory

1. Prompt template — {FLOWER_NAME} gets replaced by each flower name. Optional {SCIENTIFIC_NAME} is filled in by the import list (section 7); if omitted there, the scientific name is appended to the common name automatically.

2. Flower names — one per line

count hint

3. Batch generate parallelism

4. Generated flowers
min max

5. Crossbreed test

parent A name

parent B name

6. Botanical metadata — resolve each common name to a species and pull family, genus & native range from GBIF (backbone: Kew's WCVP), with a link to the POWO page parallelism re-enrich already-done

7. Import list — paste rows of Common name | Scientific name (markdown tables ok). Renders each image from the scientific name and pulls its metadata automatically. Uses the template in section 1.

parallelism skip names already in gallery

8. Background removal — cut each flower out onto a transparent background with PhotoRoom, saved as a copy (the original image is kept) parallelism skip images already done

S1. Seasons — one weekly roster of 10 flowers (5 free + 5 premium), served centrally to every user. Prod stores every week you push and serves the one whose week is current, rolling forward on its own. Badges are checked live against prod: on prod now is the week the app is serving, on prod ✓ is uploaded and waiting, not on prod (stale push) means the local "pushed" stamp is out of date (re-push it). current (local) only affects this machine's /api/season/current.

S2. Create / edit a season — pick exactly 10 flowers below, mark 5 as premium

week start

filter hide flowers already in a season hide flowers without a goal