website-finishing-director
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to perform multi-pass audits on external websites provided by users. This creates a surface for indirect prompt injection where a malicious website could contain instructions designed to influence the agent's audit results or behavior.
- Ingestion points: The skill instructs the agent to process content from a "live URL" provided in the user's prompt (e.g., in SKILL.md under "How to Use").
- Boundary markers: The instructions lack explicit delimiters or "ignore instructions" warnings to separate the audit framework's rules from the untrusted content being audited.
- Capability inventory: The skill encourages the agent to use tool-based access to read website content and codebase structures (mentions "codebase access" and "live URL").
- Sanitization: There are no instructions for the agent to sanitize, filter, or validate the content retrieved from target URLs before applying the audit logic.
Audit Metadata