audit-website

Pass

Audited by Gen Agent Trust Hub on Feb 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill processes data from external websites, creating an indirect prompt injection surface where a malicious site could attempt to influence the agent's behavior.
  • Ingestion points: Data is ingested from external URLs via the 'audit' command.
  • Boundary markers: The tool uses a structured hybrid XML format (, ) to separate metadata from content.
  • Capability inventory: The tool performs website analysis and reports findings; no shell write access or sensitive file read capabilities are present in the provided skill files.
  • Sanitization: Documentation in OUTPUT-FORMAT.md states all text content is XML-escaped (<, >, &) to prevent the agent from misinterpreting website content as functional instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 20, 2026, 06:05 AM