caveman

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines stylistic rules for communication to save tokens. It does not introduce any technical vulnerabilities such as command execution or network access.
  • [SAFE]: The skill demonstrates safety awareness by explicitly instructing the agent to revert to standard, clear language for security warnings and destructive action confirmations.
  • [SAFE]: The provided exit commands ('stop caveman', 'normal mode') ensure that the stylistic persona can be deactivated by the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 05:37 PM
Security Audit — agent-trust-hub — caveman