gws-modelarmor-sanitize-prompt

To sanitize user prompts, apply a Model Armor template to filter unsafe content before processing. This tool integrates into automated prompt preprocessing pipelines, ensuring organization-specific safety rules are enforced on text or JSON inputs.

Template-driven sanitization filters unsafe content from chatbot prompts by applying predefined safety rules. You can submit text or full JSON requests to produce sanitized prompts suitable for downstream AI workflows.

Yes, a valid template resource name is required to sanitize prompts with Model Armor. The template drives the sanitization process, enforcing your organization-specific safety and compliance rules on the input data.

Yes, you can sanitize a full JSON request instead of just text input. The tool accepts both --text and --json arguments to accommodate different data sources within your content moderation pipelines.

Automated prompt preprocessing should be used when integrating AI workflows that handle user-generated content. It ensures unsafe content is filtered out by a Model Armor template before the prompts reach your downstream AI applications.

Template-driven prompt sanitization matters for compliance because it enforces organization-specific safety rules on user prompts. By filtering unsafe content before processing, it ensures AI applications adhere to security and compliance requirements.