caveman-help

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a documentation-only reference. All detected patterns such as shell commands, configuration paths (~/.config/caveman/config.json), and environment variables (CAVEMAN_DEFAULT_MODE) are presented within Markdown code blocks as examples for the user. They are not instructions for the agent to execute or access.
  • [SAFE]: The external link to GitHub is a standard documentation reference to the project's source code repository and does not involve automated data exfiltration or remote code execution.
  • [SAFE]: There is no evidence of prompt injection, obfuscation, or persistence mechanisms. The skill explicitly instructs the agent to be a 'one-shot display' and not to persist any mode or state.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 09:02 PM
Security Audit — agent-trust-hub — caveman-help