Classify and filter unsafe LLM inputs and outputs across six safety categories.
npx skills add https://github.com/OpenLAIR/dr-claw --skill llamaguard-openlair
Please help me install this Agent Skill. Skill: llamaguard Source: https://github.com/OpenLAIR/dr-claw/tree/main/skills/safety-alignment/llamaguard Command: npx skills add https://github.com/OpenLAIR/dr-claw --skill llamaguard-openlair
Automatically filters and classifies unsafe user inputs and LLM outputs across six safety categories to prevent harmful interactions.
Install the required libraries and run a moderation workflow with LlamaGuard to filter unsafe prompts.
High-intent search queries and answers about installing and using this skill.
Moderate LLM interactions by classifying and filtering unsafe content across six safety categories. This skill integrates into chat apps, customer support, and API services to ensure safe user interactions and content generation.
AI content moderation covers six safety categories: violence/hate, sexual content, weapons, substances, self-harm, and criminal planning. It automatically filters unsafe user inputs and LLM outputs to prevent harmful interactions.
Yes, you can deploy LlamaGuard using vLLM, HuggingFace, and Sagemaker. It integrates with NeMo Guardrails to build production-grade moderation pipelines for API services and chat applications.
You need transformers, torch, and vllm dependencies to set up content moderation with LlamaGuard. These libraries support the moderation workflow that classifies and filters unsafe prompts and LLM outputs.
NeMo Guardrails integration supports both input and output moderation for LLMs. It filters unsafe user prompts and generated responses across six safety categories within production moderation pipelines.
The best way to filter unsafe user inputs and LLM outputs is deploying a moderation pipeline with vLLM and HuggingFace. This approach classifies content across six safety categories to prevent harmful API interactions.
lxh755818-botServe LLMs with OpenAI API, quantization, and tensor parallelism.
WaterplanAIManages multi-account Google Workspace services via a CLI with orchestration and permissions.
Clay-HHKClassifies LLM inputs and outputs into six safety categories using Meta's LlamaGuard moderation model.
ovachieverClassify user input and system outputs across six safety categories.
