Increase safety and control with pre/post hooks and built-in guardrails

You can now configure pre- and post-execution hooks for agents and enable built-in guardrails including prompt injection checks, PII detection, and OpenAI Moderation. This makes it straightforward to add safety, governance, and auditing without custom code.

Details

Run pre-execution checks and post-run audits uniformly across agents.
Block risky content and enforce policies consistently across channels and models.

Who this is for: Teams in regulated environments or those seeking stronger content governance and auditability.

‍