Skip to main content
AI. Policy Guardrails let you author boundaries the AI agents must respect, and Policy Binding attaches a policy to specific pipelines and roles. It is Upcoming, not yet available.
Policy Guardrails

​
Steps

1

Author a policy

Define the boundaries and rules the agents must follow.
Author policy
2

Review the guardrails

Confirm what the policy allows and disallows.
Review guardrails
3

Bind to a pipeline or role

Attach the policy to the pipelines and roles it should govern.
Bind policy
4

Confirm enforcement

Verify the policy is active for the bound roles.
Confirm enforcement