llamaguard

Community

Moderation with Meta's guardrails

Authorovachiever
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill introduces LlamaGuard for safe input/output moderation with high accuracy, including optional deployments via vLLM, HuggingFace, and NeMo Guardrails.

Core Features & Use Cases

  • Moderation Core: 6 safety categories (violence/hate, sexual content, weapons, substances, self-harm, criminal planning).
  • Deployment Options: vLLM, HuggingFace, and cloud/edge deployments for fast moderation.
  • API & Endpoints: FastAPI/REST style moderation endpoints with batch processing.
  • Input/Output Filtering: Pre- and post-LLM moderation to ensure safe user interactions.
  • Performance: High throughput with optimized inference.

Quick Start

Deploy LlamaGuard with a chosen backend (vLLM/HuggingFace/NeMo Guardrails) and route user input through the moderator before invoking LLMs.

Dependency Matrix

Required Modules

transformerstorchvllm

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: llamaguard
Download link: https://github.com/ovachiever/droid-tings/archive/main.zip#llamaguard

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository