shopify
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill enables the agent to fetch and process untrusted external data from Shopify stores (products, collections, metafields, and metaobjects).
- Ingestion points: External data enters the agent context through Shopify Admin API search/lookup tools and Storefront API loaders referenced in
SKILL.md. - Boundary markers: There are no explicit instructions to use delimiters or sanitization markers when the agent summarizes fetched catalog content for the user.
- Capability inventory: The skill focuses on UI generation and cart state management. It does not grant the agent capabilities to execute shell commands or perform arbitrary file writes using the ingested data.
- Sanitization: The skill describes parsing stringified JSON from metafields but lacks specific guidance on sanitizing HTML content (e.g.,
descriptionHtml) before rendering or processing. - [EXTERNAL_DOWNLOADS]: The skill references the installation of the
@shopify/hydrogen-reactpackage and the loading of a script fromcdn.shopify.comto support Shop Pay functionality. - Evidence:
references/shop-pay-button.mdspecifies loadinghttps://cdn.shopify.com/shopifycloud/shop-js/v1.0/client.js. - Context: These are official resources from Shopify, a well-known service provider, and are used for their intended primary purpose of enabling secure checkout buttons.
Audit Metadata