responsive-search-ads

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content, including landing pages and query reports, which creates a potential surface for indirect prompt injection attacks.
  • Ingestion points: Fetched landing pages, pasted export data, and competitor ad content as defined in SKILL.md and references/agent-security.md.
  • Boundary markers: The skill contains explicit instructions to treat retrieved content as data, never as instructions, and to quote any detected injection attempts rather than following them.
  • Capability inventory: The agent has the capability to browse/fetch URLs and read/write to the local .agents/product-context.md file.
  • Sanitization: Instructions include quoting injection attempts and strictly avoiding following URLs found within fetched content.
  • [PROMPT_INJECTION]: Deterministic detectors flagged instructions regarding the concealment of warnings; however, context shows these are quality-control measures to prevent users from deploying incomplete drafts, rather than malicious attempts to hide agent behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 02:23 PM