What problem does it solve? Product photos for the Armado en México catalog arrive as cutouts on opaque white backgrounds, which look broken on the site's dark gradient hero. This Skill automates the full image pipeline: background removal with alpha, edge decontamination, cropping, framing, and integration into the catalog data files for firearms, ammunition boxes, and accessories. ## Core Features & Use Cases - Background removal and edge cleanup: Uses rembg with the explicitly pinned birefnet-general model (avoiding the CC BY-NC default), decontaminates semi-transparent edges, crops to the subject, and encodes to WebP with alpha. - Automated quality control: Eleven mask and RGB metrics feed a red/amber/green semaphore, plus a generated HTML contact sheet for mandatory human review before approval. - Catalog integration: Approved images are copied into the site, cache-busting ?v= parameters are updated, and generated blocks in data.js, data-municiones.js, and data-accesorios.js are rewritten idempotently. - Use Case: When new ammunition enters the inventory, run the cajas.py cycle to find the manufacturer's box photo by brand+calibre key, process it, and apply it so every cartridge of that brand and calibre inherits the correct box image. ## Quick Start Ask the AI to prepare product photos for the pistol catalog entries and generate the contact sheet for review.
