subagent-orchestrator
Pass
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a robust architectural framework for multi-agent delegation without including malicious code, obfuscation, or unauthorized network operations.\n- [PROMPT_INJECTION]: The instructions employ persona-reinforcing language, such as 'most senior expert' and 'principled behavior'. In the context of this orchestrator, these are utilized to ensure the quality and focus of subagent output rather than to bypass security filters.\n- [COMMAND_EXECUTION]: The 'tester' role is authorized to perform command execution for verification. The skill mitigates risk by requiring subagents to stay within defined boundaries and recommending approval for risky commands.\n- [EXTERNAL_DOWNLOADS]: The README provides installation instructions via a GitHub repository belonging to the author. This is consistent with the skill's distribution and does not introduce external risk.
Audit Metadata