task-direction-approval
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [No Code] (SAFE): This skill contains no executable scripts, shell commands, or binary files. It consists entirely of natural language instructions and communication templates.
- [Safe Behavior] (SAFE): The instructions promote human-in-the-loop oversight by requiring the agent to stop and ask for permission before changing technical directions.
- [Indirect Prompt Injection] (SAFE): While the skill logic triggers based on external tool outputs (e.g., build failures), the skill lacks any capabilities (file writes, network access, or command execution) that could be exploited via malicious input. It serves purely as a reasoning constraint.
Audit Metadata