audit-fix-loop

Pass

Audited by Gen Agent Trust Hub on Apr 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a structured workflow for quality audits without malicious intent.
  • [SAFE]: Command execution is restricted to local directory initialization (mkdir) for state management.
  • [SAFE]: Potential for indirect prompt injection exists as the skill processes untrusted user artifacts, but this is inherent to its primary function and managed within subagent contexts.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 27, 2026, 08:06 AM