ask-questions-if-underspecified

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a defensive workflow that requires the agent to pause and seek clarification on ambiguous tasks, which helps prevent unintended or harmful actions.
  • [SAFE]: No evidence of malicious patterns such as prompt injection, data exfiltration, or unauthorized command execution was found in the instructions or metadata.
  • [SAFE]: The skill does not include external dependencies, scripts, or remote code patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 06:17 AM