Does this skill support PII filtering?

Yes, it includes detailed implementation patterns for detecting and masking Personally Identifiable Information (PII) using integrations like Microsoft Presidio.

Can I use this for jailbreak detection?

Absolutely. The skill provides specific workflows to define and block prompt injection attempts and jailbreak patterns before they reach your primary model.

Colang is the domain-specific language (DSL) used by NeMo Guardrails to define the conversational flows, actions, and safety rules that govern the AI's behavior.

What hardware is required for NeMo Guardrails?

While it can run on a CPU, an NVIDIA T4 GPU or better is recommended for production environments to ensure low-latency safety checks.

What is NeMo Guardrails?

NeMo Guardrails is an open-source framework by NVIDIA used to add programmable safety, topicality, and security rails to LLM-based applications at runtime.

NeMo Guardrails Safety Alignment

Name: NeMo Guardrails Safety Alignment
Author: Orchestra-Research

byOrchestra-Research

•

3,983

•

Data Science & ML

Implements programmable safety rails and runtime validation for LLM applications using NVIDIA's NeMo Guardrails framework.

This skill provides comprehensive guidance and implementation patterns for NVIDIA's NeMo Guardrails, an open-source framework for adding programmable safety rails to Large Language Model (LLM) applications. It enables developers to enforce behavioral constraints, detect jailbreak attempts, filter PII, and perform real-time fact-checking using the Colang 2.0 domain-specific language. Designed for production environments, this skill helps ensure your AI agents remain safe, on-topic, and factually accurate by wrapping LLM interactions with configurable verification layers.

Key Features

01Automated PII filtering with Presidio integration

02Programmable safety rails using Colang 2.0 DSL

033,983 GitHub stars

04Real-time jailbreak and prompt injection detection

05Hallucination detection and factual consistency checking

06Support for LlamaGuard and custom moderation models

Use Cases

01Preventing sensitive data leaks in customer-facing chatbots

02Validating factual accuracy of LLM responses in research environments

03Building production-ready AI agents with strict safety boundaries

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add orchestra-research/ai-research-skills nemo-guardrails

For use in Claude.ai and ChatGPT

Download Skill