nemotron-policy-generator

Generate BYO safety policies for NVIDIA Nemotron guardrails with Markdown, JSON, and system prompt artifacts.

Updated Apr 16, 2026
One-click install
npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill nemotron-policy-generator
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nemotron-policy-generator
Source: https://github.com/sayalinvidia/sayali-skills-test/tree/main/skills/nemotron-policy-generator
Command: npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill nemotron-policy-generator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) and references (resource) components.

What problem does it solve?

This skill helps security and product teams generate BYO custom safety policies for NVIDIA Nemotron guardrails, covering both Nemotron-Content-Safety-Reasoning-4B (text) and Nemotron-3-Content-Safety (multimodal) models, and outputs a deployable artifact suite.

Core Features & Use Cases

  • Generate BYO safety policies that map rough requirements to a structured policy artefact compatible with Nemotron guardrails.
  • Produce a canonical Markdown policy, a JSON taxonomy, and a drop-in system prompt suitable for deployment in enterprise RAG copilots, with optional custom categories and locale support.
  • Export artifacts with references to reusable templates and validation workflows (as described by SKILL.md) to aid adoption and governance.

Quick Start

Specify your policy inputs and run the generator to export a Markdown policy, a JSON taxonomy, and a Nemotron system prompt ready for deployment.

Frequently Asked Questions about nemotron-policy-generator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create custom safety policies for NVIDIA Nemotron guardrails?

You can generate custom safety policies for Nemotron guardrails by specifying your requirements to produce a deployable artifact suite containing a Markdown policy, JSON taxonomy, and system prompt.

How do I build a safety taxonomy for enterprise RAG copilots?

Building a safety taxonomy for enterprise RAG copilots involves mapping rough requirements to V2 canonical categories with optional custom categories, outputting a structured JSON taxonomy and system prompt.

Does the Nemotron policy generator support multimodal content safety models?

Yes, the policy generator supports both the text-only Nemotron-Content-Safety-Reasoning-4B and the multimodal Nemotron-3-Content-Safety models for creating guardrail policies.

What artifacts do I need to deploy Nemotron content safety guardrails?

Deploying Nemotron content safety guardrails requires a policy artifact suite consisting of a canonical Markdown policy, a JSON taxonomy, and a drop-in Nemotron system prompt.

Can I add custom categories to Nemotron safety policies?

Yes, you can add optional custom categories and per-category metadata alongside V2 canonical categories when generating Nemotron safety policies for enterprise deployments.

When should I use BYO safety policies instead of default Nemotron guardrails?

Use BYO safety policies when default guardrails do not cover specific enterprise risk-management needs, allowing you to define custom categories and metadata for governance in RAG or copilot deployments.