agent-native-audit

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate analysis of local files (configuration and source code) for the purpose of architectural auditing.
  • [SAFE]: No evidence of data exfiltration, credential harvesting, or unauthorized remote downloads was detected.
  • [SAFE]: The instructions include protective guardrails that explicitly forbid the agent from running destructive commands like 'rm' or modifying git state during the audit process.
  • [SAFE]: The skill is transparent in its operation and limited in scope to the specific project being analyzed by the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 02:20 AM
Security Audit — agent-trust-hub — agent-native-audit