amazon-marketing

name

amazon-marketing

description

Amazon product listing optimization skill for Claude Code and AI agents. Use this skill when the user mentions Amazon listings, product titles, bullet points, product descriptions, backend keywords, search terms, A+ Content, listing audit, Amazon SEO, ASIN optimization, SKU optimization, Rufus optimization, or anything related to creating, improving, or auditing Amazon product detail pages. Also trigger when the user mentions FBA/FBM product copy, Amazon keyword research, competitor listing analysis, or Brand Registry content. If the user pastes an Amazon URL, ASIN, SKU, or product data and wants copy written or improved, use this skill. Also trigger when the user wants to onboard products from `asins.csv`, create `products/[ASIN]/product-context.md`, or bootstrap product context from live Amazon pages. Also trigger when the user wants to differentiate copy across parent/child variants or customize bullets for variation children that share inherited parent content. Even if they just say "optimize my listing" or "write bullets" — this is the skill.

Amazon Marketing Skill

An open-source Claude Code skill for Amazon product listing optimization, keyword strategy, and content marketing.

What This Skill Does

Creates, optimizes, and audits Amazon product detail page content: titles, bullet points, descriptions, backend search terms, A+ Content planning, competitor analysis, and product onboarding — all grounded in Amazon's actual policies (updated through early 2026) and current best practices for both A9/A10 keyword search and Rufus AI conversational discovery.

Decision Tree

1. User wants to create a new listing from scratch? → Load skills/listing-optimizer/SKILL.md → Interview for product details using templates/listing-brief.md → Generate title → bullets → description → backend keywords

2. User wants to onboard products from ASINs/SKUs or build product context files? → Load skills/product-onboarding/SKILL.md → Read asins.csv if present → Create or update products/[ASIN]/product-context.md

3. User wants to improve an existing listing? → Load skills/listing-audit/SKILL.md first to score current state → Then load skills/listing-optimizer/SKILL.md to rewrite weak elements

4. User wants to differentiate child ASIN copy inside a variation family? → Load skills/listing-optimizer/SKILL.md → Resolve the parent product context plus all relevant child product contexts → Apply the variation-child differentiation workflow → Write per-child drafts back into each child's ## Listing Draft section

5. User wants keyword research? → Load skills/keyword-research/SKILL.md → Accept seed keywords, competitor ASINs, competitor SKUs, or product descriptions → Output structured keyword map

6. User wants competitor analysis? → Load skills/competitor-analysis/SKILL.md → Tear down 3-5 competitor listings

7. User wants A+ Content planning? → Load skills/a-plus-content/SKILL.md → Plan module layout, write copy, create image briefs

8. User wants a quick fix on one element (e.g., "just fix my title")? → Jump directly to that section in skills/listing-optimizer/SKILL.md

Context Loading

Before generating any product-specific content, resolve context in this order:

1. Load brand-context.md from the project root or .agents/ directory if it exists
2. Resolve the product the user names by SKU or ASIN, then load its products/[ASIN]/product-context.md
3. If multiple products exist and the user has not specified one, ask which SKU or ASIN to use

For variation-family requests, load the parent context and every relevant child context together before deciding which claims stay shared versus which need child-specific copy.

When resolving a product reference, sellers may identify products by SKU or ASIN. The canonical file path remains products/[ASIN]/product-context.md, so use the SKU: field inside the file for human-friendly lookup when the seller provides a SKU instead of an ASIN.

If no product context file exists, either run skills/product-onboarding/SKILL.md or interview the user for minimum viable context before writing:

• What is the product? (category, key features, materials)
• Who is it for? (target buyer, use cases)
• What makes it different? (vs. competitors)
• Are you Brand Registered? (affects A+ Content eligibility)
• What marketplace(s)? (US, CA, UK — limits vary)

When reading product context, prefer this provenance order:

1. User Confirmed Data
2. Amazon Observed Data
3. AI Inferences

Do not treat inferences as confirmed business facts.

Do not assume shared mechanics, capacities, or included items across variation children unless the product context or user has confirmed them.

Reference Files

All sub-skills should consult these before generating content:

• references/amazon-style-guide.md — Character limits, prohibited terms, formatting rules
• references/rufus-optimization.md — How to optimize for Amazon's AI shopping assistant
• references/backend-keywords-guide.md — 249-byte rules, formatting, what to include/exclude

Optional MCP Integrations

This skill works standalone with zero paid dependencies. If MCP servers are available, sub-skills can optionally use:

MCP Server	Enhancement
Playwright MCP / CLI	Browser-based PDP capture for ASIN onboarding
AmazonSeller-mcp-server	Pull live catalog data and current listing copy
Amazon Ads MCP (Official)	Tie listing changes to ad performance data
KuudoAI/amazon_ads_mcp	Campaign data for keyword prioritization
Seller Labs MCP	Profitability data to prioritize ASINs
PPC Prophet MCP	PPC data + n8n workflow integration