guardrails

Configure self-check input validation and harmful-message detection via the NeMo platform SDK.

58|16|Updated May 14, 2026
One-click install
npx skills add https://github.com/NVIDIA-NeMo/nemo-platform --skill guardrails-nvidia-nemo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guardrails
Source: https://github.com/NVIDIA-NeMo/nemo-platform/tree/main/agents/nemo-agent-local/src/nemo_agent/skills/guardrails
Command: npx skills add https://github.com/NVIDIA-NeMo/nemo-platform --skill guardrails-nvidia-nemo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the critical need for AI safety by providing a structured mechanism to intercept and evaluate user inputs against predefined security policies before they reach the model.

Core Features & Use Cases

  • Self-Check Configuration: Easily define and deploy input validation flows to ensure incoming prompts meet safety standards.
  • Harmful Message Detection: Automatically scan and block malicious or prohibited content using the NeMo platform SDK.
  • Use Case: A developer building a customer-facing chatbot can use this Skill to ensure that any input containing toxic language or prohibited topics is blocked at the gateway level, preventing the model from generating unsafe responses.

Quick Start

Use the guardrails skill to configure a self-check input flow and verify that a harmful message is correctly blocked by the platform.

Frequently Asked Questions about guardrails

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

FAQPage Schema
How do I validate AI inputs and block harmful content before it reaches the model?

Input validation intercepts and evaluates user prompts against predefined security policies. You can configure self-check input flows using the NeMo platform SDK to block malicious content at the gateway level before inference.

How do I set up a self-check input flow for an AI agent?

To set up a self-check input flow, use the guardrails skill to define validation configurations via the NeMo platform SDK. This deploys safety standards that automatically scan and block prohibited topics in incoming prompts.

Does the NeMo platform SDK support harmful-message detection for customer-facing chatbots?

Yes, the NeMo platform SDK supports harmful-message detection. It automatically scans and blocks toxic language or prohibited topics, ensuring customer-facing chatbots do not generate unsafe responses.

What's the best way to manage security policies and monitor inference traffic for AI agents?

The best way to manage security policies is by integrating with the platform's guardrail API. This monitors inference traffic and enforces input validation rules to intercept harmful content before processing.

Can I use guardrails to block toxic language at the gateway level?

Yes, you can use guardrails to block toxic language at the gateway level. The skill creates self-check configurations that evaluate incoming prompts against safety standards to prevent unsafe content from reaching the model.

Why do I need input validation to ensure AI safety?

Input validation is needed to ensure AI safety because it provides a structured mechanism to intercept and evaluate user inputs against predefined security policies before they reach the model, preventing the generation of unsafe responses.