behavior-preserving-refactor
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- Instructional Integrity: The instructions prioritize maintaining existing code behavior and warn against unauthorized changes or cleanup during structural refactors, promoting safe development practices.- Toolchain Usage: The skill recommends using repository-local toolchains and following existing build/test instructions, which is a standard and expected practice for a developer-oriented agent.- No Risky Patterns: Analysis of the skill instructions found no evidence of prompt injection, data exfiltration, obfuscation, or unauthorized remote code execution.
Audit Metadata