Install
$ agentstack add skill-commerce-agentic-agentic-commerce-skills-agentic-commerce-audit-descriptions ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Purpose
Drills into the description dimension of the Agentic Commerce Spec score. AI shopping agents do not parse marketing copy the way humans do — they extract structured facts: materials, dimensions, use cases, compatibility, target persona. Descriptions that read beautifully for humans can still be invisible to AI agents if they lack these structured signals.
This skill surfaces:
- Descriptions below the recommended word count (20+ words for AI parseability)
- Descriptions missing use-case language ("perfect for", "designed for", "works with")
- Descriptions missing measurement/dimension data
- Descriptions missing material/composition data
- Descriptions that are duplicate or near-duplicate across products
Prerequisites
- Authenticated Shopify CLI session:
shopify auth login --store - API scopes:
read_products
Parameters
| Parameter | Type | Required | Default | Description | |-----------|------|----------|---------|-------------| | store | string | yes | — | Store domain | | format | string | no | human | Output format: human or json | | limit | integer | no | 250 | Max products to audit | | minwords | integer | no | 20 | Minimum description word count to consider "AI-parseable" | | flagduplicates | bool | no | true | Flag products with near-duplicate descriptions |
Workflow Steps
- OPERATION:
products— query with description fields only (lighter than full audit)
- For each product description:
- Strip HTML, count words
- Detect use-case language via keyword pattern matching (
perfect for|designed for|ideal for|works with|compatible with|suitable for) - Detect measurements via regex for dimensions (
\d+(?:\.\d+)?\s*(?:cm|mm|in|inches|kg|lb|oz)) - Detect materials via vertical-specific keyword lists (e.g., apparel: cotton, wool, polyester; home: oak, stainless steel)
- Flag duplicates if
flag_duplicates: true: compute SHA-256 of description text → group products by hash. Surface groups of 2+ products sharing identical descriptions.
- Output ranked report:
- Critical: descriptions Read-only operation. No mutations.
GraphQL Operations
# products:query — focused on description fields
query DescriptionAudit($first: Int!, $after: String) {
products(first: $first, after: $after) {
edges {
node {
id
title
descriptionHtml
productType
vendor
}
}
pageInfo {
hasNextPage
endCursor
}
}
}
Companion fix
After this audit, the [agentic-commerce-generate-ai-descriptions](../../catalog-fix/agentic-commerce-generate-ai-descriptions/SKILL.md) skill can auto-rewrite flagged descriptions, preserving brand voice while injecting the missing structured signals.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: commerce-agentic
- Source: commerce-agentic/agentic-commerce-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.