Генерация и редактирование изображений через gpt-image-1

★ 7.6 · media-processing

gpt-image is a Claude Code skill that generates and edits images using the gpt-image-2 model (OpenAI), with support for text-to-image, reference-image editing, and inpainting via alpha mask. The skill bundles a 162-prompt Reference Gallery Atlas organized by category, plus a craft.md checklist for dense text, diagrams, UI mockups, data visualizations, and multi-panel layouts. All execution goes through the `gpt-image` CLI or `scripts/generate.py` — no ad-hoc generation code is written. Output options include PNG, JPEG, and WebP at sizes from 1k to 4k, three quality tiers, and background contro