seo-audit
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external domains during the audit process.
- Ingestion points: Website content is fetched via
run_site_audit,get_audit_pages, and manual HTML retrieval in the verification step (Step 5). - Boundary markers: The skill does not implement explicit delimiters or 'ignore' markers to isolate the ingested site content from its core instructions.
- Capability inventory: The agent can modify project context (
update_project_context,addKeyPages,appendResearchLog) and output results via theseo-reporttool. - Sanitization: A mandatory 'adversarial pass' (Step 8) involves a self-review or second-agent review to verify findings against facts and filter out drama or unglossed jargon, providing a mitigation layer against malicious inputs.
Audit Metadata