菜单
Use when the user asks how a category or the overall assortment is performing, where the margin and volume sit, which products or categories are strong or weak, or where there is headroom to grow margin. Produces an evidence-backed portfolio view across categories and personas.
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
Operations runbook for the agentic-supply-chain repo. USE THIS SKILL when the user asks to deploy, build, provision, ingest, index, register, or clean up any part of the project — including infrastructure (azd), container images, Container Apps, AI Search indexes, Foundry toolboxes, and hosted agents. Covers the full lifecycle: provision → build → index → ingest → deploy → register → clean up.
Use when the user asks to plan a promotional campaign, build a weekly flyer, decide which products to discount, or design a multi-week promotion calendar. Produces a structured, margin-aware campaign plan grounded in competitor promotions and internal pricing.
Use when the user asks to optimise the price of a product or category, find the margin-maximising price, decide how deep a discount can go, or respond to a specific competitor price. Produces an elasticity- and margin-grounded pricing recommendation.
| name | portfolio-analysis |
| description | Use when the user asks how a category or the overall assortment is performing, where the margin and volume sit, which products or categories are strong or weak, or where there is headroom to grow margin. Produces an evidence-backed portfolio view across categories and personas. |
Analyse the product portfolio across categories and personas to find where margin and volume concentrate and where there is headroom to act.
list_categories for the aggregate weekly
margin, volume and average margin % per category. Rank categories by margin
contribution and by margin %.get_category_margin_forecast and list_products to see
the per-SKU spread. Identify margin leaders, margin laggards and high-volume
traffic drivers.list_personas and map each category to the
personas that shop it most (category_affinity). Note where a high-margin
category depends on a price-sensitive persona (fragile) vs a low-sensitivity
persona (defensible).search_competitor_promotions to see which of
our strong categories competitors are actively attacking.Return a category-level table: category, weekly volume, weekly margin €, avg margin %, dominant persona, competitive pressure (low/med/high), and classification. Follow with 3–5 prioritised recommendations, each tied to the evidence (which tool/number drove it).