designing-leakage-safe-experiments

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured guidelines and templates (Markdown files) to help users plan machine learning experiments without introducing data leakage. It does not contain any executable scripts or tool invocations.
  • [SAFE]: No external network calls, downloads, or remote code execution patterns were detected. All references are to local files within the skill package.
  • [SAFE]: The documentation follows best practices for data science and does not attempt to bypass safety filters or perform prompt injection.
  • [SAFE]: No obfuscated content, hidden URLs, or suspicious metadata were identified during the analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 01:39 PM
Security Audit — agent-trust-hub — designing-leakage-safe-experiments