catalog-content-audit-ecommerce
Pass
Audited by Gen Agent Trust Hub on Nov 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local script,
../scripts/catalog_scan.py, when handling catalogs larger than 150 SKUs to ensure accuracy in auditing. This is a functional requirement for the skill's stated purpose but involves local code execution. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided catalog exports which may contain untrusted third-party content (e.g., supplier descriptions), creating a surface for indirect prompt injection.
- Ingestion points: Catalog export files containing product titles, descriptions, and attributes (referenced in SKILL.md).
- Boundary markers: No specific delimiters or "ignore" instructions are provided to separate the untrusted catalog content from the agent's system instructions.
- Capability inventory: The skill possesses file-read capabilities for catalog exports and execution capabilities for the
catalog_scan.pyscript. - Sanitization: The instructions do not specify any sanitization or validation steps for the product data before it is analyzed or passed to the processing script.
Audit Metadata