skills.rest
tadod12
tadod12COMMUNITY
Explore all skills by this author →
GitHub

llamaguard

Classify and filter unsafe user and model content with LlamaGuard-7b.

Updated Mar 18, 2026
Search Tags:#moderation#guardrails#production-deployment#content-safety#llamaguard#prompt-filtering
One-click install
npx skills add https://github.com/tadod12/fraud-detection-research --skill llamaguard-tadod12
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: llamaguard
Source: https://github.com/tadod12/fraud-detection-research/tree/main/.agent/skills/07-safety-alignment/llamaguard
Command: npx skills add https://github.com/tadod12/fraud-detection-research --skill llamaguard-tadod12

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

LlamaGuard provides specialized moderation for LLM inputs and outputs, ensuring safer conversations by automatically classifying and filtering unsafe content across multiple categories.

Core Features & Use Cases

  • Category-based safety classification across violence, sexual content, weapons, substances, self-harm, and criminal planning.
  • Integration-ready workflows: input filtering, output moderation, vLLM deployment, API serving, and NeMo Guardrails.
  • Real-world use: deploy in chat apps to prevent unsafe prompts and responses.

Quick Start

Install transformers and torch, login to HuggingFace, load the LlamaGuard-7b model, and run moderation on user messages to filter unsafe content.

Frequently Asked Questions about llamaguard

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

FAQPage Schema
How do I filter unsafe LLM prompts before they reach the model?▼

To filter unsafe LLM prompts, you can apply input filtering using LlamaGuard-7b to classify text across categories like violence and self-harm, blocking inappropriate messages before processing.

Can I use LlamaGuard with NeMo Guardrails for content moderation?▼

Yes, LlamaGuard integrates directly with NeMo Guardrails to enforce content safety, providing category-based classification and filtering for both user inputs and model outputs.

How do I deploy LlamaGuard for API serving in production?▼

You can deploy LlamaGuard for production API serving using vLLM, loading the LlamaGuard-7b model from the HuggingFace repository to moderate chat app conversations at scale.

What are the prerequisites to run LlamaGuard locally?▼

Running LlamaGuard locally requires installing the transformers and torch stack, authenticating with the HuggingFace model repository to download LlamaGuard-7b, and configuring your local environment.

How does LlamaGuard classify unsafe content categories?▼

LlamaGuard classifies unsafe content by evaluating text against specific categories including violence, sexual content, weapons, substances, self-harm, and criminal planning to enforce safety.

Does LlamaGuard support output moderation for LLM responses?▼

Yes, LlamaGuard supports output moderation by evaluating model-generated responses against its safety categories, filtering unsafe LLM outputs before they are returned to the user.

Related Skills

Explore More Software Engineering Skills→
JeffallanJeffallan
10.9k

fine-tuning-expert

Automate fine-tuning workflows for large language models using LoRA/QLoRA and PEFT methods.

Community
Advanced
Supporter09Supporter09

llamaguard

Classify LLM chat messages into six safety categories for moderation.

Community
Advanced
arsityarsity

llamaguard

Classify user-generated content into six safety categories for LLM moderation.

Community
Intermediate
ricablericable

llamaguard

Classify LLM inputs and outputs into six safety categories.

Community
Advanced