NVIDIA/skills

Agent Skills for NVIDIA products — install into Claude Code, Codex, and other coding agents to run Physical AI, robotics, simulation, CUDA, and RAG workflows end to end.

What it solves

This project provides a catalog of portable instruction sets, called "skills," that teach AI coding agents (such as Claude Code, Codex, Cursor, and others) how to optimally use NVIDIA's vast ecosystem of software, libraries, and hardware tools. It bridges the gap between general-purpose AI agents and the specialized technical knowledge required to navigate NVIDIA's complex developer tools, from CUDA-X libraries to robotics and simulation workflows.

How it works

Skills are defined as portable instruction sets that can be installed into an AI agent via a CLI tool. The repository acts as a central mirror and catalog of these skills, which are maintained in their respective product repositories and synced daily. Users can install specific skills or browse the catalog using the npx skills command, targeting specific agents like Claude Code or Cursor.

Who it’s for

Developers and researchers using AI coding agents who need to implement NVIDIA-specific workflows, such as GPU-accelerated dataframes (cuDF), numerical optimization (cuOpt), biomolecular research (BioNeMo), or Jetson device management.

Highlights

  • Broad Ecosystem Support: Covers a wide range of NVIDIA products including CUDA-Q, cuDF, cuOpt, BioNeMo, DeepStream, and Jetson BSP.
  • Agent-Agnostic Specification: Designed to work across multiple AI agents including Claude Code, Codex, Snowflake CoCo, Cursor, and Kiro.
  • Automated Sync: Skills are mirrored daily from product repositories to ensure the latest guidance is available.
  • Detailed Catalog: Provides a comprehensive list of skills for diverse domains like medical AI, digital health, and physical AI/robotics.

Related

  • Project
    NVIDIA-BioNeMo/bionemo-agent-toolkit

    The NVIDIA BioNeMo Agent Toolkit provides ready‑to‑install “skills” that let LLM agents (Claude‑Code, Codex, etc.) call NVIDIA’s life‑science models and tools—covering protein design, structure prediction, virtual screening, genomics, and cheminformatics. Skills are installed via the `skills` CLI or native plugin marketplaces, include scripts and step‑by‑step instructions, and come with an evaluation harness to measure how much they improve an agent’s performance.

  • Project
    onmax/nuxt-skills

    A collection of specialized skills for AI coding agents that provides deep, context-aware guidance for the Vue, Nuxt, and Nitro development ecosystems.

  • Project
    waynesutton/builder-skills

    builder‑skills provides a set of concise, markdown‑based “skills” that teach AI coding assistants how to follow Convex best‑practice rules and disciplined project‑workflow habits (PRDs, task tracking, safe git). Installable via skills.sh, Claude plugins, or npm, it drops starter docs and lets agents load the appropriate SKILL.md on demand, keeping code, docs, and safety checks in sync.

  • Project
    amd/skills

    AMD Skills is a catalog of reusable, agent‑compatible “skills” that encode AMD‑specific AI workflow knowledge (ROCm diagnostics, Quark quantisation, LLM serving on Instinct/EPYC, etc.). Install via the `skills` CLI (`npx skills add amd/skills`) or manually, then invoke the skills through coding assistants like Cursor, Claude Code, Codex or Gemini. Each skill bundles a `SKILL.md` description, optional scripts and references, enabling agents to automatically perform setup, optimisation and debugging steps for AMD hardware.

  • Project
    dzhng/skills

    A library of reusable, composable AI‑agent “skills” (prompt‑driven actions) for building autonomous software‑factory loops, installable via npm and usable with many coding assistants.