s4h-investigation-evidence-audit

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured reasoning instructions for performing evidence audits. It includes human-in-the-loop checkpoints via the AskUserQuestion tool to confirm framing and scope before proceeding with analysis.
  • [SAFE]: No malicious patterns such as prompt injection, obfuscation, or data exfiltration attempts were detected. The instructions focus on classification, quality assessment, and conflict-of-interest checks.
  • [SAFE]: There are no external dependencies, remote script downloads, or dynamic code execution patterns. References to other modules (e.g., /s4h-investigation-source-trace) are presented as workflow suggestions for the user rather than automated execution triggers.
  • [SAFE]: The skill does not request access to sensitive files, environment variables, or credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 11:44 PM
Security Audit — agent-trust-hub — s4h-investigation-evidence-audit