← ClaudeAtlas

photo-croplisted

Calculates correct object-position for HTML/CSS photos in containers with object-fit: cover. Reads each photo via Read tool, maps subject position and size, analyzes container-to-image ratio, derives X/Y. Flags when crop is too tight and suggests container proportion fix. Use when: "установи object-position", "настрой кроп фото", "photo crop", "object-position для фото", "фото обрезается неправильно", "set object-position", "fix photo crop", "портрет обрезает голову", "настрой позицию фото", "кадрирование фото"
stepanenkoviktor0110-boop/ai-dev-methodology · ★ 1 · AI & Automation · score 57
Install: claude install-skill stepanenkoviktor0110-boop/ai-dev-methodology
# Photo Crop — object-position Calculator Part of: Design Pipeline (use after frontend-design:frontend-design when placing photos in layouts) Apply this workflow for every `<img>` with `object-fit: cover`. Each photo is unique — run all four steps individually per photo without reusing values from other photos. ## Shot size vocabulary Before placing any photo, decide the **target shot size** for the container context: | Shot size (RU) | Shot size (EN) | Shows | |---|---|---| | Дальний / Общий план | Extreme Long / Very Long Shot | Full environment, tiny person | | Полный план | Long Shot | Full body head to feet | | Поколенный план | Medium Long Shot | Head to knees | | Средний план | Medium Shot | Head to waist | | Полукрупный план | Medium Close Up | Head to chest/shoulders | | Крупный план | Close Up | Face + neck | | Предельно крупный | Big Close Up | Face only | **Default targets by section type:** - Hero (full-screen): **Medium Shot** or **Medium Long Shot** — show enough body to read the outfit - About / portrait column: **Medium Close Up** — face + shoulders readable - Guide / work section: **Medium Shot** — person + environment context **CRITICAL rule:** Defaulting Y to show just the face = Close Up. This is almost always wrong for hero sections. Always ask: "what shot size does this section need?" before setting Y. ## Step 1 — Read the photo Use Read tool to view the image. Identify and note: - **Subject**: face, body, key object — what must stay visible -