content-diff
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches HTML content from external Adobe-managed domains (*.aem.page and *.aem.live) to perform content comparisons, which is the primary intended function.
- [COMMAND_EXECUTION]: Instructions suggest using curl or equivalent tools to fetch raw HTML, ensuring that metadata and block attributes are preserved for an accurate author-friendly report.
- [SAFE]: The skill explicitly mitigates indirect prompt injection risks in the 'External Content Safety' section. Evidence: Ingestion occurs via external HTML fetching (SKILL.md Step 2); Boundary markers are established in the safety guidelines; Capability inventory is limited to content retrieval and text-based reporting; Sanitization is implemented through instructions to treat all fetched data as untrusted and to avoid executing scripts.
Audit Metadata