# Visual credits

All images and the video in `public/visuals/` were generated on 27 October 2026 with Higgsfield AI (https://higgsfield.ai), using the models `gpt_image_2_5` (hero still), `cinematic_studio_2_5` (section, transform and category stills) and Kling 3.0 (via Higgsfield) for the hero loop `hero-loop.mp4`.

- They are illustrative only. They do not depict real products, real businesses, real people or real brands, and they carry no logos or readable text.
- The category images (`cat-*.webp`) are generic stand-ins for store heroes when a scraped page has no usable image. They must not be presented as the photos of any specific business.
- Post-processing: resized and converted to WebP (quality 80); the video was re-encoded to H.264 MP4 with ffmpeg.
- Prompts and job IDs are recorded in the team's project journal, outside this repository.