Language: JapaneseEnglish

amd-skills Claude Code Plugin

Author
AMD
Category
Development
Topics
AI Agents & AI App Development · Claude Code Customization & Workflow · Monitoring & Observability
First cataloged
2026-07-17 (UTC)
Explanation last updated
2026-07-17 (UTC)
Source (GitHub) last updated
2026-07-16 (UTC) (yesterday)

The explanation below is AI-generated. Please verify it against the sources.

amd-skills is a plugin that bundles a set of AMD-authored Agent Skills for working with AMD hardware and software. According to the README, it covers routing image, text-to-speech and speech-to-text tasks through local AI on Ryzen AI systems, serving large language models on AMD Instinct GPUs and EPYC CPUs with vLLM, and analyzing GPU kernel and PyTorch trace performance. Each skill packages instructions, helper scripts and reference material that an agent loads only when a matching task comes up. Per the README, some listed skills are already implemented in the repository while others are marked as planned, so the catalog is still expanding.

What you can do with amd-skills

  • Wire local AI into an existing cloud LLM application for offline support, improved privacy, and lower API costs (local-ai-app-integration skill)
  • Deploy and benchmark LLM inference on AMD Instinct GPUs with vLLM, including hardware detection and recipe selection (serving-llms-on-instinct skill)
  • Serve LLMs on AMD EPYC CPUs using vLLM and zentorch inside a container or conda environment, with model-fit and hardware checks (serving-llms-on-epyc skill)
  • Evaluate GPU kernel correctness and performance and compare kernel implementations (magpie-kernel-evaluator skill)
  • Orchestrate PyTorch profiler trace analysis and produce a prioritized performance report (tracelens-analysis-orchestrator skill)

Sources

History of amd-skills

  • Claude Code Plugin Updated Repository of amd-skills
  • Claude Code Plugin Added amd-skills

Back to list