search-page-audit
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and parses content from external web pages, including HTML source code, robots.txt, sitemap.xml, and llms.txt files. This untrusted data could potentially contain malicious instructions designed to influence the agent's behavior during the audit process.
- Ingestion points: Content is retrieved from user-provided URLs as defined in the 'Fetch & Parse' section of SKILL.md.
- Boundary markers: The instructions lack explicit delimiters or warnings to treat the fetched content as data rather than instructions, which is a common surface for indirect injection.
- Capability inventory: The agent has internet access to read content and the ability to generate detailed text reports.
- Sanitization: No specific sanitization or filtering logic is described to handle potentially malicious instructions embedded in the target website's metadata or text.
- [EXTERNAL_DOWNLOADS]: The skill requires network access to fetch the audit targets. It specifically looks for standard files like robots.txt, sitemap.xml, and the llms.txt standard on the target domain to evaluate AI crawler directives and site structure.
Audit Metadata