gws-modelarmor-sanitize-response

Sanitize model outputs by applying a Model Armor template to model-generated text or JSON before delivery. This enforces policy-compliant formatting and content, reducing the risk of unsafe or non-compliant responses in automation pipelines.

A Model Armor template defines the safety and compliance policies applied to sanitize model outputs. It acts as a configuration layer that enforces policy-compliant formatting and content filtering before a response reaches the end user.

Run the gws modelarmor sanitize-response command with your specified template to sanitize text or JSON inputs. This CLI-first workflow integrates quickly into existing automation pipelines to filter outbound model responses.

Yes, the gws CLI is required to apply the Model Armor template for sanitization. You also need a valid Model Armor template configured to process your outbound text or JSON model responses.

Yes, Model Armor sanitization is designed for outbound chat responses in customer support and data-to-text generation. It ensures that model outputs are sanitized and safe for delivery in critical safety workflows.