agent-enforcement

To enforce a secure agent whitelist for multi-agent orchestration, you use a PreToolUse hook that intercepts Task or Bash tool calls and validates agent selections against a predefined whitelist before execution proceeds. This ensures deterministic CLI calls for external models and reliable instruction following for internal models.

Multi-agent Task commands fail when referencing insecure paths because secure path validation mechanisms block temporary directories like /tmp/ in Task prompts to prevent orchestration violations and ensure reliable debugging. Activating enforcement mechanisms diagnoses the issue and prevents unexpected failures.

A PreToolUse hook for agent rule enforcement is a mechanism that intercepts and blocks invalid Task or Bash tool calls, ensuring that multi-agent orchestration adheres to predefined security rules and agent selections before execution. It guarantees deterministic CLI calls and reliable instruction following.

You block invalid agent selections during team orchestration by activating enforcement mechanisms that validate agent choices against a secure whitelist, leveraging model selection to ensure better instruction following in complex scenarios. This operates via a PreToolUse hook to intercept and block invalid calls.

Yes, you can achieve reliable instruction following for internal models in complex orchestration scenarios by leveraging model selection, which ensures deterministic CLI calls for external models and validates agent selections against a whitelist. This guarantees reliable multi-agent orchestration and debugging.

The limitation of using temporary paths like /tmp/ in Task prompts is that secure path validation will actively prevent their use, blocking the Task execution to avoid insecure directory references during multi-agent orchestration. This ensures reliable multi-agent orchestration and debugging failures.