Sanitize user prompts with configurable safety rules via the gws CLI.
npx skills add https://github.com/73ai/openbotkit --skill gws-modelarmor-73ai
Please help me install this Agent Skill. Skill: gws-modelarmor Source: https://github.com/73ai/openbotkit/tree/main/skills/gws-modelarmor Command: npx skills add https://github.com/73ai/openbotkit --skill gws-modelarmor-73ai
Google Model Armor helps ensure safe, policy-compliant prompts by filtering and sanitizing user input before it reaches the AI model, reducing the risk of unsafe or leaking content.
Run the gws modelarmor command to sanitize a sample prompt before execution.
High-intent search queries and answers about installing and using this skill.
Prompt sanitization applies predefined safety templates to user inputs, removing or neutralizing unsafe content before the AI model processes the request. This ensures policy-compliant workflows in chatbots and developer environments.
Prompt sanitization is the process of filtering sensitive or unsafe content from user inputs before they reach an AI model. You need it in chatbots and assistants to enforce content policies and reduce the risk of data leakage.
Yes, prompt sanitization can operate on-device for local processing. This keeps data on your machine and preserves privacy by applying safety rules before model execution.
You can configure safety rules by applying predefined safety templates through the gws CLI. Running the gws modelarmor command sanitizes sample prompts by enforcing these rules before generation.
Yes, predefined safety templates are required to enforce content policies in chatbots. These templates define the rules for filtering and neutralizing unsafe content before model calls.
Local prompt sanitization requires configurable safety rules and integration via the gws CLI. It depends on predefined templates to filter content, meaning unconfigured rules may allow unsafe prompts to bypass the safety checks.
niclasmartinsson-sysCreate a Google Model Armor template via the gws CLI with project, location, and template-id.
jongwonyRun Gemini prompts via CLI with model selection and approval modes.
73aiSanitize user prompts through a Model Armor template with the gws tool.
RazisafirValidate LLM inputs and outputs with NeMo Guardrails and Python validators.
