delegating-to-agents

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill explicitly recommends running agents with permission-bypassing/untrusted flags (e.g., "claude --print --permission-mode bypassPermissions" and "codex --yolo"), which encourages bypassing sandboxing or host permission controls.