best-of-n
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements strong architectural security patterns for AI agent operations. By utilizing isolated worktrees for parallel execution, it prevents unintended side effects or race conditions in the parent checkout. The requirement for a 'PASS' verdict from automated verification tools (tests/lint) and an independent judging phase ensures that generated content is validated before integration. There are no instances of credential exposure, suspicious network activity, or obfuscated code detected in the skill's instructions.
Audit Metadata