finding-changelogs

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill defines a process for ingesting untrusted data from external URLs (changelogs, RSS feeds, and markdown files), which represents an indirect prompt injection surface. Mitigation is provided through explicit verification instructions.\n
  • Ingestion points: External websites and repositories accessed via WebFetch and RSS discovery logic (SKILL.md).\n
  • Boundary markers: The skill includes a 'Content Verification' section with specific criteria (red flags) to ensure the agent evaluates the relevance and safety of the ingested content.\n
  • Capability inventory: The instructions involve the use of tools such as manage_source, manage_product, and get_catalog_entry to store and organize retrieved data.\n
  • Sanitization: The skill describes automatic noise reduction for specific providers (e.g., stripping Fern MDX attributes).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 11:56 PM
Security Audit — agent-trust-hub — finding-changelogs