cro-optimization
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is configured to fetch and process external landing page content which is inherently untrusted and could contain hidden malicious instructions (e.g., in HTML comments or invisible tags) designed to influence the agent's audit behavior.
- Ingestion points: External landing page content retrieved via the WebFetch tool and raw HTML/CSS provided in the $ARGUMENTS field or follow-up messages.
- Boundary markers: The instructions lack explicit directives for the agent to treat fetched content strictly as data or to ignore any embedded instructions within that content.
- Capability inventory: The skill uses the Read tool for local reference files and WebFetch for retrieving external web content, but does not utilize file-write or system-level execution tools.
- Sanitization: There is no requirement or implementation of sanitization or validation for the external content before it is interpolated into the analysis context.
Audit Metadata