behavior-preservation-checker

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, providing methodologies for differential testing and behavior comparison without executing or downloading untrusted content.
  • [SAFE]: No evidence of obfuscation, data exfiltration, credential harvesting, or prompt injection was detected in the skill content or metadata.
  • [SAFE]: The Python code examples provided are pedagogical templates and do not contain patterns associated with remote code execution or privilege escalation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 09:19 PM
Security Audit — agent-trust-hub — behavior-preservation-checker