commercetools-data

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection vulnerability identified in 'references/bulk-catalog-enrichment.md'.
  • Ingestion points: Product attributes, names, and category data from the commercetools catalog entering the agent context via 'iterateAllProducts'.
  • Boundary markers: Absent; product data is directly interpolated into the LLM prompt template.
  • Capability inventory: The skill demonstrates calling the Anthropic API to generate product descriptions and SEO metadata.
  • Sanitization: No sanitization or escaping is applied to the catalog data before it is sent to the LLM, creating a surface where malicious product data could influence the generated output.
  • [EXTERNAL_DOWNLOADS]: The skill references several official SDKs and client libraries.
  • Details: Uses well-known packages including '@commercetools/platform-sdk', '@anthropic-ai/sdk', '@commercetools/ts-client', and '@commercetools/importapi-sdk' from official registries. These are documented as safe vendor resources.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 03:56 PM