do-and-judge
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The instruction file establishes an orchestrator role that delegates tasks cleanly via
Tasktool abstraction layers to specialized sub-agents. - [SAFE]: No commands for arbitrary execution, persistent mechanisms, data harvesting, or remote payload fetching are present in the provided instruction logic.
Audit Metadata