geo-auditor
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process content from user-provided public websites, which presents a surface for indirect prompt injection if an audited site contains malicious instructions intended to influence the agent's behavior.
- Ingestion points: The audit workflow samples content from homepages, product pages, about pages, robots.txt, and sitemaps (SKILL.md, references/audit-checklist.md).
- Boundary markers: The skill emphasizes evidence labeling (e.g., "Observed", "Indexed/cached") but lacks explicit instructions or delimiters to isolate the agent from potentially malicious instructions embedded in the external data.
- Capability inventory: The skill restricts the agent to read-only actions and does not provide capabilities for code execution, file modification, or network operations beyond standard site browsing.
- Sanitization: There is no mention of sanitizing or filtering external content before analysis.
Audit Metadata