ask-questions-if-underspecified

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists exclusively of markdown-based instructions for a clarification workflow. It does not include any scripts, commands, or executable logic.
  • [SAFE]: There are no remote downloads or external network references. All templates and examples are provided as static text.
  • [SAFE]: The instructions do not attempt to bypass safety filters or perform prompt injection; instead, they enforce a 'pause before acting' policy which reduces the risk of unintended actions.
  • [SAFE]: No sensitive file paths or credentials are referenced or accessed.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 08:11 AM