← ClaudeAtlas

product-shotslisted

Front-door router for the product-shots ecosystem. Clarifies underspecified visual creation requests through a 4-stage state machine, injects platform visual DNA (Amazon + 7 social platforms) and industry visual DNA (7 industries), enforces a unified Negative Constraints prompt patch, then routes to one of five downstream business skills (product-shots-main-image / product-shots-detail-page / product-shots-multi-angle / product-shots-ad-creative / product-shots-social-post). Use when the user says "I need a product image", "design something for my listing", "I need content for Instagram", "make an ad for me", "make a cover image", "create a post", "做一个商品图", "帮我做个详情页", "帮我做个广告图", "做一个社媒图", "做一张图", "帮我设计", "做个封面" — i.e. any underspecified visual creation request that needs clarification before generation. This is the intent-routing hub of the product-shots ecosystem.
motiful/product-shots · ★ 0 · AI & Automation · score 75
Install: claude install-skill motiful/product-shots
# Hub Expert design guidance that clarifies user intent through structured questions and produces platform-ready creative briefs enriched with visual DNA, prompt patches, and industry-specific style rules. Acts as the **intent router** of the product-shots skill ecosystem: it owns the 0–5-round clarification loop, locks the Brief, and dispatches the work to one of five downstream business skills (`product-shots-main-image` / `product-shots-detail-page` / `product-shots-multi-angle` / `product-shots-ad-creative` / `product-shots-social-post`). ## Engagement Principles These rules apply across every Section. Read before acting. 1. **Hard cap on clarification** — at most 5 rounds, terminate early when `brief_completeness > 0.8`. Never loop indefinitely. 2. **Match the user's language** — auto-detect from user input; respond in the same language for ALL responses; NEVER switch languages proactively. 3. **Ask in priority order** — questions are asked in order of "how much it changes the creative direction": Platform > Style preference > Variant count. High-impact questions first. 4. **One dimension per question** — each round asks one and only one dimension. NEVER chain several questions in one turn. 5. **Use `<suggestion>` for option sets** — every offered option is wrapped in `<suggestion>` tags so the UI renders clickable chips. Custom user input is still allowed in parallel. 6. **Inject visual DNA before routing** — every Brief carries the platform prompt patch + industry