subagent
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill functions primarily as an operational protocol to manage subagent behavior and ensure task isolation.
- [SAFE]: Explicitly instructs the agent to escalate security-sensitive changes, such as authentication, secrets, and input validation, to the orchestrator rather than performing them autonomously.
- [COMMAND_EXECUTION]: Includes standard development environment commands like 'git branch --show-current', 'mkdir', and common quality gate commands (npm run check, cargo test, go build). These are used for environment verification and quality assurance within the defined subagent scope.
- [PROMPT_INJECTION]: No patterns of instruction override, safety filter bypass, or system prompt extraction were found. The instructions reinforce existing safety boundaries and role-based access control.
Audit Metadata