nemo-guardrails

Detect jailbreaks, validate inputs, filter PII, and check toxicity in LLM applications.

6|3|Updated Jan 29, 2026
One-click install
npx skills add https://github.com/jonnabio/ace-framework --skill nemo-guardrails-jonnabio
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nemo-guardrails
Source: https://github.com/jonnabio/ace-framework/tree/main/.ace/packs/ai-research/nemo-guardrails
Command: npx skills add https://github.com/jonnabio/ace-framework --skill nemo-guardrails-jonnabio

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires nemoguardrails, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the critical need for runtime safety in LLM applications, providing features like jailbreak detection, input/output validation, fact-checking, PII filtering, and toxicity detection.

Core Features & Use Cases

  • Jailbreak Detection: Identifies and prevents unauthorized access attempts.
  • Input/Output Validation: Ensures data integrity and prevents malformed inputs.
  • Fact-Checking: Validates factual claims made by the LLM.
  • PII Filtering: Protects sensitive user information.
  • Toxicity Detection: Identifies and mitigates harmful content.
  • Use Case: For a company developing an LLM chatbot, this Skill can be used to ensure that the chatbot adheres to safety guidelines and doesn't inadvertently provide harmful or illegal advice.

Quick Start

Install the nemo-guardrails package and configure it to enforce safety rules for your LLM application.

Frequently Asked Questions about nemo-guardrails

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

FAQPage Schema
How do I implement LLM safety for jailbreak detection and toxicity filtering in production?

Runtime LLM safety for jailbreak detection and toxicity filtering is implemented by installing the nemo-guardrails package and configuring safety rules to validate inputs and outputs. It requires Colang 2.0 DSL for defining specific safety guidelines.

What is PII filtering and how does it protect sensitive user information in LLM chatbots?

PII filtering protects sensitive user information by intercepting and validating LLM inputs and outputs to prevent unauthorized data disclosure. It enforces data integrity and ensures production chatbots adhere to privacy safety guidelines.

Can I use nemo-guardrails for fact-checking and input validation without a GPU?

Fact-checking and input validation via nemo-guardrails can run without a dedicated GPU, but an NVIDIA T4 GPU is required for optimal performance. The Skill is designed for production use with LLMs where throughput matters.

What's the best way to programmatically enforce safety rules for an LLM application?

Programmatically enforcing LLM safety rules is best handled by defining rules in Colang 2.0 DSL and applying them through the nemo-guardrails package. This approach provides runtime jailbreak detection and toxicity mitigation.

How does runtime security for LLMs handle unauthorized access attempts?

Runtime security for LLMs handles unauthorized access attempts through jailbreak detection, which identifies and prevents prompt injection attacks. It validates interactions against defined Colang 2.0 rules to block malicious inputs.

Do I need Colang 2.0 DSL to configure input and output validation for my chatbot?

Yes, Colang 2.0 DSL is required to define the specific input and output validation rules for your chatbot. The nemo-guardrails package uses these rules to enforce data integrity and prevent malformed inputs.