菜单
Extract brand/product evidence from an ecommerce product URL before any Brand Shoot Kit planning or generation. Use for PDP scouting, source evidence capture, product image URL collection, degraded-mode notes, and scout.json creation. Do not use for ad spying or competitor ad research.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Turn a product URL or product image into a brand-aware ecommerce shoot plan: brand read, visual gap audit, shoot strategy, shot list, GPT Image-ready prompts, QA rubric, and export map. Use for PDP/lifestyle/model/seasonal product photography systems. Do not use for ad spying, competitor ad intelligence, paid-social creative mining, or media-buying analysis.
Package approved Brand Shoot Kit assets into channel-specific ecommerce export folders and manifests. Use for PDP/social/email/marketplace maps, naming, export manifests, and handoff-ready asset organization.
Write reusable Brand Shoot Kit memory after a product shoot: product preservation lessons, winning visual directions, failed prompts, QA risks, and future-shot guidance.
Create the immutable product truth and preservation brief for Brand Shoot Kit outputs. Use when product shape, packaging, label text, claims, variants, compliance, or distortion risk must be protected before image generation.
Write GPT Image-ready ecommerce product photography prompts from a Brand Shoot Kit shot list and product preservation brief. Use for prompt packs, negative constraints, reroll instructions, and category-aware generation guidance.
Score generated Brand Shoot Kit images for product accuracy, realism, brand fit, commerce usefulness, and artifact risk. Use for QA reports, reject queues, reroll decisions, and correction prompts.
| name | brand-scout |
| description | Extract brand/product evidence from an ecommerce product URL before any Brand Shoot Kit planning or generation. Use for PDP scouting, source evidence capture, product image URL collection, degraded-mode notes, and scout.json creation. Do not use for ad spying or competitor ad research. |
| metadata | {"openclaw":{"emoji":"🔎","user-invocable":true,"requires":{"bins":["bash","python3"],"env":[]}}} |
Capture evidence before strategy. Your job is not to be creative yet; it is to prevent hallucinated creative direction.
Produce scout.json with:
url, title, meta_description, og_title, og_descriptionh1[], image_urls[]product_name, brand_name, product_type, product_category, price,
variants, claims_benefits, ingredients_materials_specs,
visible_packaging_text_candidates, image_evidence, field_confidence,
extraction_warningsdegraded_mode and notescrape_provenance (object) recording which adapter ran:
{scraper, scraper_version, scraped_at, request_url, fixture_used, forced_by, firecrawl_meta?}.Executable module owner path:
scripts/modules/brand_scout.py --url <product-url> --packet <packet-dir> -> scout.jsonscripts/scout-structured.py --in <scout.json> --out <scout.json> for structured enrichment only.Scraper selection on the module:
--scraper {auto,curl,firecrawl} (default auto).auto follows env precedence: BSK_FORCE_SCRAPER (used by ./evals/run.py) → FIRECRAWL_API_KEY → curl default.--scraper curl recovery hint.A downstream agent should be able to cite where each product, tone, visual, and preservation clue came from.