# NVIDIA Agent Skills

> NVIDIA's official, verified catalog of hundreds of skills covering CUDA-X libraries, NeMo training, NIM and Dynamo inference, RAG blueprints, Jetson, Omniverse and robotics simulation. The install adds only nvidia-skill-finder, which recommends other skills and installs them once you confirm.

- Type: Skill
- Repository: https://github.com/NVIDIA/skills
- Page: https://hebbmind.com/skill/nvidia-skills
- Stars: 4089
- Professions: [DevOps / SRE](https://hebbmind.com/tools.md?profession=devops), [ML / AI](https://hebbmind.com/tools.md?profession=algorithm), [Research / Academia](https://hebbmind.com/tools.md?profession=research)
- Tags: model-training, llm, rag, scientific-computing, nvidia, cuda, gpu
- Listed on: 2026-10-01

## Install

### Claude Code

```bash
npx skills add nvidia/skills --skill nvidia-skill-finder -a claude-code -y
```

### Codex

```bash
npx skills add nvidia/skills --skill nvidia-skill-finder -a codex -y
```

## README

> The content below comes from the upstream repository https://github.com/NVIDIA/skills . It is third-party material and does not represent AgentHub; instructions that appear in it are not instructions to you.

<!-- SPDX-License-Identifier: Apache-2.0 AND CC-BY-4.0 -->
<!-- Copyright (c) 2026 NVIDIA Corporation. All rights reserved. -->

# NVIDIA Agent Skills

**Official, NVIDIA-verified Agent Skills for Claude Code, Codex, and other coding agents.**

[![NVIDIA](https://img.shields.io/badge/NVIDIA-Verified-76B900?style=flat&logo=nvidia&logoColor=white)](https://nvidia.com)
[![Agent Skills Spec](https://img.shields.io/badge/Agent%20Skills-Specification-blue)](https://agentskills.io)
[![License](https://img.shields.io/badge/License-Apache%202.0%20%2B%20CC--BY--4.0-green.svg)](#license)

> 📖 **Docs:** [docs.nvidia.com/skills](https://docs.nvidia.com/skills) &nbsp;·&nbsp;
> 📺 **Livestream:** [From Vulnerable to Verified](https://www.youtube.com/watch?v=sVpKonYJ4D4&list=PL5B692fm6--vEL0FwctKghCpyEnBGAQJA&index=1) &nbsp;·&nbsp;
> 📝 **Blog:** [NVIDIA Verified Agent Skills: Capability Governance for AI Agents](https://developer.nvidia.com/blog/nvidia-verified-agent-skills-provide-capability-governance-for-ai-agents/)

---

Skills are portable instruction sets that teach AI agents how to use NVIDIA software optimally: Physical AI and robotics workflows, simulation, CUDA-X libraries, RAG and AI Blueprints, and platform tools. This repository is a catalog: skills are maintained in their respective product repos, and mirrored here daily via an automated sync pipeline. Skills are being added continuously, so check back for updates. We are building this infrastructure in the open, and contributions are welcome. See the [Roadmap](#roadmap) for what is planned next.

---

## Quickstart

Install NVIDIA skills with the default [`skills` CLI](https://github.com/vercel-labs/skills) flow. The NVIDIA Skill Finder skill will search this catalog and help you discover the right skill for your use case.

```bash
npx skills add nvidia/skills --skill nvidia-skill-finder --yes
```

This installs the skill finder, which can then find and install other skills for you. To browse the full catalog instead, run `npx skills add nvidia/skills`. You do not need to clone this repo or copy skill folders by hand.

> **Requires a current `skills` CLI (v1.5.16 or newer).** Installing via `npx skills@latest add nvidia/skills` always uses the latest. On older CLIs (v1.5.15 and earlier), skills may install but not appear in Claude Code — see [Troubleshooting](https://github.com/NVIDIA/skills/blob/HEAD/docs/advanced-install.mdx#troubleshooting).

The skill is available the next time your agent loads skills and encounters a relevant task. For example, ask your agent to "solve a linear programming problem with cuOpt" and the skill guides it through the cuOpt Python API. In Claude Code, run `/reload-skills` to load newly installed skills in your current session.

### Install One Skill Without Prompts

Use this when you already know the skill name and want to skip prompts.

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --yes
```

Replace `cuopt-numerical-optimization-api` with any skill name from the [Skill Catalog](#skill-catalog).

### Install for a Specific Agent

Use `--agent` to target a specific AI coding agent. Initially, we'll support common client targets, expanding the list over time. For the full list of clients supported by the spec, see the [`skills` CLI Supported Agents table](https://github.com/vercel-labs/skills#supported-agents).

**Claude Code**

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent claude-code
```

**Codex**

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent codex
```

**Snowflake CoCo**

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent cortex
```

**Cursor**

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent cursor
```

**Kiro**

```bash
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent kiro-cli
```

Use `--agent` more than once to install the same skill into multiple agents.

```bash
npx skills add nvidia/skills \
  --skill cuopt-numerical-optimization-api \
  --agent claude-code \
  --agent codex \
  --agent cursor \
  --agent kiro-cli
```

### Keep Skills Up to Date

New skills land continuously, and existing ones are revised, renamed, or consolidated as the catalog evolves. Refresh what you have installed with:

```bash
npx skills update
```

Run it interactively and the CLI also flags skills that were removed or merged upstream (for example, when several skills are consolidated into one) and offers to remove the stale local copies. Use `npx skills list` to see what is installed and `npx skills check` to preview what is out of date first.

### Browse the Catalog

Use this when you want to see available NVIDIA skills before installing anything.

```bash
npx skills add nvidia/skills --list
```

For non-interactive installs, global installs, agent-specific installs, updates, removals, and fallback manual copying, see [Advanced installation](https://github.com/NVIDIA/skills/blob/HEAD/docs/advanced-install.mdx).

---

## Skill Catalog

<!-- skills-table-start -->
| Product | Description | Skills |
|---------|-------------|--------|
| **ALCHEMI Toolkit** | GPU-accelerated atomistic simulation and machine-learned interatomic potential workflows, including batched relaxation, molecular dynamics, training and fine-tuning, and Zarr datasets. | [`nvalchemi-toolkit`](skills/nvalchemi-toolkit) |
| **Auto Ontology** | Set up Auto Ontology, model and publish its semantic layer, and run validated text-to-SQL queries through it. | [`nvidia-ontology-setup`](skills/nvidia-ontology-setup), [`nvidia-ontology-management`](skills/nvidia-ontology-management), [`nvidia-ontology-query`](skills/nvidia-ontology-query) |
| **BioNeMo Libraries** | Accelerate cheminformatics, genomic analysis, biomolecular structure prediction, and equivariant model development with NVIDIA GPU libraries and tools. | [`bioir-bench-perf-oss`](skills/bioir-bench-perf-oss), [`bioir-make-data-pipeline`](skills/bioir-make-data-pipeline), [`bioir-model-onboard`](skills/bioir-model-onboard), [`bionemo-cuequivariance`](skills/bionemo-cuequivariance), [`bionemo-cuequivariance-jax`](skills/bionemo-cuequivariance-jax), [`bionemo-cuequivariance-torch`](skills/bionemo-cuequivariance-torch), [`bionemo-nvmolkit-usage`](skills/bionemo-nvmolkit-usage), [`bionemo-parabricks`](skills/bionemo-parabricks), [`bionemo-genomics-workflow-acceleration`](skills/bionemo-genomics-workflow-acceleration) |
| **BioNeMo NIMs** | Run and combine BioNeMo NIM microservices for biomolecular research and drug discovery through hosted APIs or local deployments. | [`bionemo-drug-discovery-pipeline`](skills/bionemo-drug-discovery-pipeline), [`bionemo-msa-structure-prediction-pipeline`](skills/bionemo-msa-structure-prediction-pipeline), [`bionemo-openfold2-nim`](skills/bionemo-openfold2-nim), [`bionemo-molmim-nim`](skills/bionemo-molmim-nim), [`bionemo-diffdock-nim`](skills/bionemo-diffdock-nim), [`bionemo-genmol-nim`](skills/bionemo-genmol-nim), [`bionemo-evo2-nim`](skills/bionemo-evo2-nim), [`bionemo-rfdiffusion-nim`](skills/bionemo-rfdiffusion-nim), [`bionemo-rfd4-proteina-nim`](skills/bionemo-rfd4-proteina-nim), [`bionemo-openfold3-nim`](skills/bionemo-openfold3-nim), [`bionemo-proteinmpnn-nim`](skills/bionemo-proteinmpnn-nim), [`bionemo-msa-search-nim`](skills/bionemo-msa-search-nim), [`bionemo-boltz2-nim`](skills/bionemo-boltz2-nim) |
| **BioNeMo Open Models** | Apply BioNeMo open models to molecular property prediction and generative protein design. | [`bionemo-codonfm-setup`](skills/bionemo-codonfm-setup), [`bionemo-codonfm-embed`](skills/bionemo-codonfm-embed), [`bionemo-codonfm-score`](skills/bionemo-codonfm-score), [`bionemo-codonfm-finetune`](skills/bionemo-codonfm-finetune), [`bionemo-complexa-evaluate-pdbs`](skills/bionemo-complexa-evaluate-pdbs), [`bionemo-complexa-sweep`](skills/bionemo-complexa-sweep), [`bionemo-complexa-target`](skills/bionemo-complexa-target), [`bionemo-kermt-add-cmim-pretrain`](skills/bionemo-kermt-add-cmim-pretrain), [`bionemo-kermt-continue-pretrain`](skills/bionemo-kermt-continue-pretrain), [`bionemo-kermt-embed`](skills/bionemo-kermt-embed), [`bionemo-kermt-finetune`](skills/bionemo-kermt-finetune), [`bionemo-kermt-infer`](skills/bionemo-kermt-infer), [`bionemo-kermt-monitor`](skills/bionemo-kermt-monitor), [`bionemo-kermt-pretrain-scratch`](skills/bionemo-kermt-pretrain-scratch), [`bionemo-kermt-setup`](skills/bionemo-kermt-setup) |
| **CDF AnomalyGen** | Run Cosmos Data Factory AnomalyGen workflows for fine-tuning, synthetic defect generation, evaluation, refinement, and pseudo-labeling. | [`cdf-anomalygen`](skills/cdf-anomalygen) |
| **CDF Augmentation** | Author, validate, and run Cosmos Data Factory augmentation pipelines for image and video generation and transformation. | [`cdf-augmentation`](skills/cdf-augmentation) |
| **CDF Auto-Labeling** | Build and run Cosmos Data Factory auto-labeling pipelines for video enhancement, tracking, captioning, and question generation. | [`cdf-auto-labeling`](skills/cdf-auto-labeling) |
| **CDF Curation and Retrieval** | Author, validate, and run Cosmos Data Factory curation and retrieval pipelines. | [`cdf-curation-and-retrieval`](skills/cdf-curation-and-retrieval) |
| **CDF Simulation** | Build and validate Cosmos Data Factory Simulation jobs and compiled payloads for PCB-AOI workflows. | [`cdf-simulation`](skills/cdf-simulation) |
| **CompileIQ** | Compiler autotuning for CUDA C++, Triton, and Helion kernels; set up CompileIQ, apply Booster Packs, pick a search space, write the objective, run and debug the search, and validate the winning configuration. | [`compileiq-author-objective`](skills/compileiq-author-objective), [`compileiq-booster-pack`](skills/compileiq-booster-pack), [`compileiq-bootstrap`](skills/compileiq-bootstrap), [`compileiq-debug`](skills/compileiq-debug), [`compileiq-run-search`](skills/compileiq-run-search), [`compileiq-search-space`](skills/compileiq-search-space), [`compileiq-validate-result`](skills/compileiq-validate-result) |
| **Cosmos Data Factory Orchestration** | Build, run and monitor Cosmos Data Factory pipelines for image and video generation and transformation. | [`cdf-orchestration-preflight`](skills/cdf-orchestration-preflight), [`cdf-orchestration-setup`](skills/cdf-orchestration-setup), [`cdf-orchestration-write-dag`](skills/cdf-orchestration-write-dag), [`cdf-autonomous-vehicle-dataset-generation`](skills/cdf-autonomous-vehicle-dataset-generation), [`cdf-defect-image-generation`](skills/cdf-defect-image-generation), [`cdf-event-video-generation`](skills/cdf-event-video-generation), [`cdf-image-attribute-augmentation`](skills/cdf-image-attribute-augmentation), [`cdf-video-data-augmentation`](skills/cdf-video-data-augmentation) |
| **CUDA-Q** | Use for CUDA-Q setup, simulation targets, QPU access, and @cudaq.kernel authoring guidance. | [`cudaq-guide`](skills/cudaq-guide), [`cudaq-importing`](skills/cudaq-importing) |
| **cuDF** | Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV/Parquet I/O, nullable semantics, and multi-GPU DataFrame workloads. | [`accelerated-computing-cudf`](skills/accelerated-computing-cudf), [`cudf-query-engine-builder`](skills/cudf-query-engine-builder) |
| **cuML** | Official NVIDIA-authored guidance for cuML GPU machine learning, scikit-learn acceleration with cuml.accel, API compatibility, multi-GPU process execution, and model export. | [`cuml`](skills/cuml) |
| **cuOpt** | GPU-accelerated optimization — vehicle routing, linear programming, quadratic programming, installation, server deployment, and developer tools. | [`cuopt-install`](skills/cuopt-install), [`cuopt-multi-objective-exploration`](skills/cuopt-multi-objective-exploration), [`cuopt-numerical-optimization-api`](skills/cuopt-numerical-optimization-api), [`cuopt-numerical-optimization-formulation`](skills/cuopt-numerical-optimization-formulation), [`cuopt-routing-api-python`](skills/cuopt-routing-api-python), [`cuopt-server-api-python`](skills/cuopt-server-api-python) |
| **DALI** | GPU-accelerated data loading and processing with NVIDIA DALI. | [`dali-dynamic-mode`](skills/dali-dynamic-mode) |
| **DeepStream** | Agentic skills for guided DeepStream development. | [`amc-run-rtsp-calibration`](skills/amc-run-rtsp-calibration), [`amc-run-sample-calibration`](skills/amc-run-sample-calibration), [`amc-run-video-calibration`](skills/amc-run-video-calibration), [`amc-setup-calibration-stack`](skills/amc-setup-calibration-stack), [`deepstream-dev`](skills/deepstream-dev), [`deepstream-generate-pipeline`](skills/deepstream-generate-pipeline), [`deepstream-import-vision-model`](skills/deepstream-import-vision-model), [`deepstream-profile-pipeline`](skills/deepstream-profile-pipeline), [`deepstream-run-mv3dt`](skills/deepstream-run-mv3dt), [`deepstream-sop`](skills/deepstream-sop), [`rtvi-cv-customize-model`](skills/rtvi-cv-customize-model), [`rtvi-cv-scaffold-vss-service`](skills/rtvi-cv-scaffold-vss-service), [`rtvi-vlm-customize-model`](skills/rtvi-vlm-customize-model) |
| **DGX Spark Skills** | Agent guidance for DGX Spark and related GB10 system identification, troubleshooting, updates, playbooks, NVIDIA Sync, diagnostics, and support. | [`dgx-spark-skills`](skills/dgx-spark-skills) |
| **Digital Health** | Agent skills for (1) the clinical ASR evaluation flywheel — term curation, synthetic clinical-speech benchmark generation, KER (Keyword Error Rate) scoring, and fine-tune guidance; (2) the ambient healthcare agent creation with Nemotron Voice Agent. | [`digital-health-clinical-asr-setup`](skills/digital-health-clinical-asr-setup), [`digital-health-clinical-asr-build`](skills/digital-health-clinical-asr-build), [`digital-health-clinical-asr-eval`](skills/digital-health-clinical-asr-eval), [`digital-health-clinical-asr-finetune`](skills/digital-health-clinical-asr-finetune), [`ambient-healthcare-agent-with-nemotron-voice-agent`](skills/ambient-healthcare-agent-with-nemotron-voice-agent) |
| **DOCA** | Teach AI agents to use the NVIDIA DOCA SDK on BlueField DPUs and ConnectX NICs — setup, libraries, services, tools, deployment, and debugging. | [`doca-bare-metal-deployment`](skills/doca-bare-metal-deployment), [`doca-bf3-deployment`](skills/doca-bf3-deployment), [`doca-bf4-deployment`](skills/doca-bf4-deployment), [`doca-container-deployment`](skills/doca-container-deployment), [`doca-debug`](skills/doca-debug), [`doca-hardware-safety`](skills/doca-hardware-safety), [`doca-programming-guide`](skills/doca-programming-guide), [`doca-public-knowledge-map`](skills/doca-public-knowledge-map), [`doca-setup`](skills/doca-setup), [`doca-structured-tools-contract`](skills/doca-structured-tools-contract), [`doca-upgrade`](skills/doca-upgrade), [`doca-version`](skills/doca-version), [`doca-aes-gcm`](skills/doca-aes-gcm), [`doca-argp`](skills/doca-argp), [`doca-comch`](skills/doca-comch), [`doca-common`](skills/doca-common), [`doca-compress`](skills/doca-compress), [`doca-devemu`](skills/doca-devemu), [`doca-dma`](skills/doca-dma), [`doca-dpa`](skills/doca-dpa), [`doca-dpdk-bridge`](skills/doca-dpdk-bridge), [`doca-erasure-coding`](skills/doca-erasure-coding), [`doca-eth`](skills/doca-eth), [`doca-flow`](skills/doca-flow), [`doca-gpunetio`](skills/doca-gpunetio), [`doca-mgmt`](skills/doca-mgmt), [`doca-pcc`](skills/doca-pcc), [`doca-rdma`](skills/doca-rdma), [`doca-rmax`](skills/doca-rmax), [`doca-sha`](skills/doca-sha), [`doca-sta`](skills/doca-sta), [`doca-telemetry`](skills/doca-telemetry), [`doca-telemetry-exporter`](skills/doca-telemetry-exporter), [`doca-urom`](skills/doca-urom), [`doca-verbs`](skills/doca-verbs), [`doca-argus`](skills/doca-argus), [`doca-dms`](skills/doca-dms), [`doca-dts`](skills/doca-dts), [`doca-firefly`](skills/doca-firefly), [`doca-urom-svc`](skills/doca-urom-svc), [`doca-bench`](skills/doca-bench), [`doca-caps`](skills/doca-caps), [`doca-comm-channel-admin`](skills/doca-comm-channel-admin), [`doca-flow-tune`](skills/doca-flow-tune), [`doca-ngauge`](skills/doca-ngauge), [`doca-pcc-counters`](skills/doca-pcc-counters), [`doca-perftest`](skills/doca-perftest), [`doca-pipeline-language-developer-tools`](skills/doca-pipeline-language-developer-tools), [`doca-sha-offload-engine`](skills/doca-sha-offload-engine), [`doca-socket-relay`](skills/doca-socket-relay), [`doca-spcx-cc`](skills/doca-spcx-cc), [`doca-telemetry-traceback-application`](skills/doca-telemetry-traceback-application), [`doca-telemetry-utils`](skills/doca-telemetry-utils) |
| **Dynamo** | NVIDIA Dynamo deployment bring-up on Kubernetes — pick and deploy recipes, start router modes, validate disagg NIXL/UCX/NCCL interconnect, and triage day-2 failures. | [`dynamo-interconnect-check`](skills/dynamo-interconnect-check), [`dynamo-recipe-runner`](skills/dynamo-recipe-runner), [`dynamo-router-starter`](skills/dynamo-router-starter), [`dynamo-troubleshoot`](skills/dynamo-troubleshoot) |
| **Earth2Studio** | Open-source deep-learning framework for exploring, building and deploying AI weather/climate workflows. | [`earth2studio-create-datasource`](skills/earth2studio-create-datasource), [`earth2studio-create-diagnostic`](skills/earth2studio-create-diagnostic), [`earth2studio-create-prognostic`](skills/earth2studio-create-prognostic), [`earth2studio-data-fetch`](skills/earth2studio-data-fetch), [`earth2studio-deterministic-forecast`](skills/earth2studio-deterministic-forecast), [`earth2studio-discover`](skills/earth2studio-discover), [`earth2studio-install`](skills/earth2studio-install) |
| **ETHOS.md** | Capture an AI agent's purpose, boundaries, and success criteria in a portable ethos.md that developers, reviewers, and evaluators share. | [`ethos-orchestrate`](skills/ethos-orchestrate), [`ethos-explore`](skills/ethos-explore), [`ethos-write`](skills/ethos-write) |
| **FastGen** | Set up FastGen, prepare datasets, run inference and diffusion distillation, and extend networks, modalities, and methods. | [`fastgen-add-method`](skills/fastgen-add-method), [`fastgen-add-modality`](skills/fastgen-add-modality), [`fastgen-add-network`](skills/fastgen-add-network), [`fastgen-configure`](skills/fastgen-configure), [`fastgen-dataset`](skills/fastgen-dataset), [`fastgen-develop`](skills/fastgen-develop), [`fastgen-inference`](skills/fastgen-inference), [`fastgen-runtime`](skills/fastgen-runtime), [`fastgen-setup`](skills/fastgen-setup), [`fastgen-train`](skills/fastgen-train), [`fastgen-validate`](skills/fastgen-validate) |
| **FoundationPose Perception Pipeline** | Set up the FoundationPose perception stack, adapt BOP datasets, and run or evaluate pose inference with TAO Deploy depth. | [`foundationpose-setup`](skills/foundationpose-setup), [`foundationpose-pipeline`](skills/foundationpose-pipeline) |
| **G-Assist** | Check or change a machine's NVIDIA display and GPU settings such as resolution, refresh rate, V-Sync, G-SYNC, brightness, color, and GPU performance, via the G-Assist MCP server. | [`g-assist-mcp-skill`](skills/g-assist-mcp-skill) |
| **HoloHub** | Build, run, debug, benchmark, and develop Holoscan applications and Holoscan Modules with validated lifecycle workflows. | [`holoscan-app-debug-build-run`](skills/holoscan-app-debug-build-run), [`holoscan-module-lifecycle`](skills/holoscan-module-lifecycle) |
| **Holoscan SDK** | Install and set up the Holoscan SDK on any platform (container, Debian, Python, Conda, or source). | [`holoscan-install-debian`](skills/holoscan-install-debian), [`holoscan-install-source`](skills/holoscan-install-source), [`holoscan-install-wheel`](skills/holoscan-install-wheel), [`holoscan-install-conda`](skills/holoscan-install-conda), [`holoscan-install-container`](skills/holoscan-install-container), [`holoscan-setup`](skills/holoscan-setup) |
| **Holoscan Sensor Bridge** | Agent-ready skills for Holoscan Sensor Bridge devkit workflows, including demo environment bring-up, FPGA flashing for Lattice and VB1940 hardware, example application execution, QA test-plan automation, and support for configuring and using the Holoscan Sensor Bridge FPGA intellectual property (IP) core. | [`hsb-setup`](skills/hsb-setup), [`hsb-flash`](skills/hsb-flash), [`hsb-app`](skills/hsb-app), [`hsb-test`](skills/hsb-test), [`hsb-ip-def`](skills/hsb-ip-def), [`hsb-ip-packetizer`](skills/hsb-ip-packetizer), [`hsb-ip-create-top`](skills/hsb-ip-create-top) |
| **Isaac for Healthcare Workflows** | Agent-ready skills for Isaac for Healthcare agentic and catheter-navigation workflows, covering task authoring, data pipelines, policy training and validation, CT-derived digital twins, DRR rendering, and interactive catheter simulation. | [`i4h-workflow`](skills/i4h-workflow), [`i4h-workflow-setup`](skills/i4h-workflow-setup), [`i4h-workflow-create`](skills/i4h-workflow-create), [`i4h-workflow-scene-edit`](skills/i4h-workflow-scene-edit), [`i4h-workflow-dataset-teleop`](skills/i4h-workflow-dataset-teleop), [`i4h-workflow-dataset-replay`](skills/i4h-workflow-dataset-replay), [`i4h-workflow-dataset-mimic`](skills/i4h-workflow-dataset-mimic), [`i4h-workflow-dataset-annotate`](skills/i4h-workflow-dataset-annotate), [`i4h-workflow-dataset-convert`](skills/i4h-workflow-dataset-convert), [`i4h-workflow-finetune`](skills/i4h-workflow-finetune), [`i4h-workflow-train-rl`](skills/i4h-workflow-train-rl), [`i4h-workflow-validate`](skills/i4h-workflow-validate), [`i4h-workflow-e2e`](skills/i4h-workflow-e2e), [`i4h-lerobot-viz`](skills/i4h-lerobot-viz) |
| **Ising** | Connect to the NVIDIA Ising Calibration model through the hosted API or a self-hosted NIM, with credential setup and connectivity checks. | [`ising-calibration-setup`](skills/ising-calibration-setup) |
| **Jetson BSP** | Agentic skills for setting up and customizing an NVIDIA Jetson Linux Board Support Package (BSP) — pick a target, prepare image and sources, customize IO (camera, PCIe, USB, pinmux, clocks, and more), then promote, flash, and validate. | [`jetson-build-source`](skills/jetson-build-source), [`jetson-customize-camera`](skills/jetson-customize-camera), [`jetson-customize-clocks`](skills/jetson-customize-clocks), [`jetson-customize-fan`](skills/jetson-customize-fan), [`jetson-customize-mgbe`](skills/jetson-customize-mgbe), [`jetson-customize-nvpmodel`](skills/jetson-customize-nvpmodel), [`jetson-customize-pcie`](skills/jetson-customize-pcie), [`jetson-customize-pinmux`](skills/jetson-customize-pinmux), [`jetson-customize-uphy`](skills/jetson-customize-uphy), [`jetson-customize-usb`](skills/jetson-customize-usb), [`jetson-derive-carrier`](skills/jetson-derive-carrier), [`jetson-download-bsp`](skills/jetson-download-bsp), [`jetson-flash-image`](skills/jetson-flash-image), [`jetson-generate-kb`](skills/jetson-generate-kb), [`jetson-init-image`](skills/jetson-init-image), [`jetson-init-source`](skills/jetson-init-source), [`jetson-init-target`](skills/jetson-init-target), [`jetson-link-docs`](skills/jetson-link-docs), [`jetson-optimize-memory`](skills/jetson-optimize-memory), [`jetson-print-bsp-info`](skills/jetson-print-bsp-info), [`jetson-promote-image`](skills/jetson-promote-image), [`jetson-quick-start`](skills/jetson-quick-start), [`jetson-set-target`](skills/jetson-set-target), [`jetson-validate-image`](skills/jetson-validate-image) |
| **Jetson Device** | Device-side agent skills for working with a live NVIDIA Jetson after boot — diagnostics, memory auditing, headless setup, inference memory tuning, LLM serving and benchmarking, packaging guidance, and speculative decoding. | [`jetson-diagnostic`](skills/jetson-diagnostic), [`jetson-headless-mode`](skills/jetson-headless-mode), [`jetson-inference-mem-tune`](skills/jetson-inference-mem-tune), [`jetson-llm-benchmark`](skills/jetson-llm-benchmark), [`jetson-llm-serve`](skills/jetson-llm-serve), [`jetson-memory-audit`](skills/jetson-memory-audit), [`jetson-package`](skills/jetson-package), [`jetson-print-device-info`](skills/jetson-print-device-info), [`jetson-speculative-decoding`](skills/jetson-speculative-decoding), [`jetson-video-benchmark`](skills/jetson-video-benchmark), [`jetson-video-capability`](skills/jetson-video-capability), [`jetson-video-pipeline`](skills/jetson-video-pipeline), [`jetson-video-recipe`](skills/jetson-video-recipe), [`jetson-video-setup`](skills/jetson-video-setup) |
| **Kumo Engine** | Build, train and run inference with Kumo relational models from a coding agent. | [`kumo-data-science`](skills/kumo-data-science), [`kumoai-reference`](skills/kumoai-reference), [`kumo-model-tuning`](skills/kumo-model-tuning), [`kumo-model-persistence`](skills/kumo-model-persistence), [`kumo-model-explanations`](skills/kumo-model-explanations), [`kumo-training-diagnostics`](skills/kumo-training-diagnostics), [`kumo-custom-populations`](skills/kumo-custom-populations), [`kumo-graph-visualization`](skills/kumo-graph-visualization) |
| **Medical AI Skills** | Agent-ready medical AI skills built on MONAI for DICOM handling, NVIDIA-hosted medical imaging model workflows, segmentation, synthesis, and evidence-oriented evaluation. | [`dicom-metadata-extract`](skills/dicom-metadata-extract), [`dicom-series-preflight`](skills/dicom-series-preflight), [`dicom-series-to-volume`](skills/dicom-series-to-volume), [`medtech-model-evidence-export`](skills/medtech-model-evidence-export), [`nv-generate-ct-rflow`](skills/nv-generate-ct-rflow), [`nv-generate-mr`](skills/nv-generate-mr), [`nv-generate-mr-brain`](skills/nv-generate-mr-brain), [`nv-generate-mr-brain-finetune`](skills/nv-generate-mr-brain-finetune), [`nv-generate-vae-finetune`](skills/nv-generate-vae-finetune), [`nv-reason-ct`](skills/nv-reason-ct), [`nv-reason-cxr`](skills/nv-reason-cxr), [`nv-segment-ct`](skills/nv-segment-ct), [`nv-segment-ct-finetune`](skills/nv-segment-ct-finetune), [`nv-segment-ctmr`](skills/nv-segment-ctmr) |
| **Megatron-Core** | Large-scale distributed training — model parallelism, pipeline parallelism, and mixed precision. | [`mcore-create-issue`](skills/mcore-create-issue), [`mcore-linting-and-formatting`](skills/mcore-linting-and-formatting), [`mcore-run-on-slurm`](skills/mcore-run-on-slurm), [`mcore-split-pr`](skills/mcore-split-pr), [`mcore-testing`](skills/mcore-testing) |
| **NeMo AutoModel** | NeMo AutoModel - PyTorch-native distributed training for LLMs/VLMs with Hugging Face support, recipes, launchers, and validation workflows. | [`nemo-automodel-distributed-training`](skills/nemo-automodel-distributed-training), [`nemo-automodel-launcher-config`](skills/nemo-automodel-launcher-config), [`nemo-automodel-model-onboarding`](skills/nemo-automodel-model-onboarding), [`nemo-automodel-recipe-development`](skills/nemo-automodel-recipe-development) |
| **NeMo Data Designer** | Build declarative synthetic dataset generation pipelines with NeMo Data Designer. | [`data-designer`](skills/data-designer) |
| **NeMo Fabric** | Portable skills for integrating applications with NeMo Fabric through its public SDK and building compatible third-party adapters. | [`nemo-fabric-integrate`](skills/nemo-fabric-integrate), [`nemo-fabric-build-adapter`](skills/nemo-fabric-build-adapter) |
| **NeMo MBridge** | NeMo MBridge - PyTorch-native bridge between Hugging Face and Megatron-Core for checkpoint conversion, training recipes, and NVIDIA GPU performance workflows. | [`nemo-mbridge-mlm-bridge-training`](skills/nemo-mbridge-mlm-bridge-training), [`nemo-mbridge-multi-node-slurm`](skills/nemo-mbridge-multi-node-slurm), [`nemo-mbridge-perf-activation-recompute`](skills/nemo-mbridge-perf-activation-recompute), [`nemo-mbridge-perf-cpu-offloading`](skills/nemo-mbridge-perf-cpu-offloading), [`nemo-mbridge-perf-cuda-graphs`](skills/nemo-mbridge-perf-cuda-graphs), [`nemo-mbridge-perf-expert-parallel-overlap`](skills/nemo-mbridge-perf-expert-parallel-overlap), [`nemo-mbridge-perf-hierarchical-context-parallel`](skills/nemo-mbridge-perf-hierarchical-context-parallel), [`nemo-mbridge-perf-megatron-fsdp`](skills/nemo-mbridge-perf-megatron-fsdp), [`nemo-mbridge-perf-memory-tuning`](skills/nemo-mbridge-perf-memory-tuning), [`nemo-mbridge-perf-moe-comm-overlap`](skills/nemo-mbridge-perf-moe-comm-overlap), [`nemo-mbridge-perf-moe-dispatcher-selection`](skills/nemo-mbridge-perf-moe-dispatcher-selection), [`nemo-mbridge-perf-moe-hardware-configs`](skills/nemo-mbridge-perf-moe-hardware-configs), [`nemo-mbridge-perf-moe-long-context`](skills/nemo-mbridge-perf-moe-long-context), [`nemo-mbridge-perf-moe-optimization-workflow`](skills/nemo-mbridge-perf-moe-optimization-workflow), [`nemo-mbridge-perf-moe-vlm-training`](skills/nemo-mbridge-perf-moe-vlm-training), [`nemo-mbridge-perf-parallelism-strategies`](skills/nemo-mbridge-perf-parallelism-strategies), [`nemo-mbridge-perf-sequence-packing`](skills/nemo-mbridge-perf-sequence-packing), [`nemo-mbridge-perf-tp-dp-comm-overlap`](skills/nemo-mbridge-perf-tp-dp-comm-overlap), [`nemo-mbridge-recipe-recommender`](skills/nemo-mbridge-recipe-recommender), [`nemo-mbridge-resiliency`](skills/nemo-mbridge-resiliency) |
| **NeMo Relay** | Skills to help get started and use NeMo Relay - a runtime for instrumenting and controlling AI agents across harnesses, applications, and frameworks. | [`nemo-relay-install`](skills/nemo-relay-install), [`nemo-relay-get-started`](skills/nemo-relay-get-started), [`nemo-relay-instrument-calls`](skills/nemo-relay-instrument-calls), [`nemo-relay-instrument-context-isolation`](skills/nemo-relay-instrument-context-isolation), [`nemo-relay-instrument-typed-wrappers`](skills/nemo-relay-instrument-typed-wrappers), [`nemo-relay-plugin-adaptive-tuning`](skills/nemo-relay-plugin-adaptive-tuning), [`nemo-relay-plugin-build`](skills/nemo-relay-plugin-build), [`nemo-relay-plugin-observability`](skills/nemo-relay-plugin-observability), [`nemo-relay-migrate-from-flow`](skills/nemo-relay-migrate-from-flow), [`nemo-relay-debug-runtime-integration`](skills/nemo-relay-debug-runtime-integration) |
| **NeMo Retriever** | NeMo Retriever - deploy NeMo Retriever Library locally, extract information from corpus of data, and answer questions against the corpus. | [`nemo-retriever`](skills/nemo-retriever), [`nemo-retriever-mcp`](skills/nemo-retriever-mcp) |
| **NeMo-RL** | RLHF training on Ray — GRPO, DPO, and SFT for LLMs and VLMs with FSDP2 and Megatron-Core. | [`launch-nemo-rl`](skills/launch-nemo-rl), [`nemo-rl-auto-research`](skills/nemo-rl-auto-research), [`nemo-rl-brev-etiquette`](skills/nemo-rl-brev-etiquette), [`nemo-rl-docs`](skills/nemo-rl-docs), [`nemo-rl-session-memory`](skills/nemo-rl-session-memory) |
| **NemoClaw** | Secure agent sandboxing — run OpenClaw inside NVIDIA OpenShell with managed inference, policy management, remote deployment, sandbox monitoring. | [`nemoclaw-user-guide`](skills/nemoclaw-user-guide) |
| **Nemotron** | Author end-to-end model development, customization, evaluation, and deployment pipelines using the NVIDIA AI stack. | [`nemotron-customize`](skills/nemotron-customize), [`nemotron-retrieval-recipes`](skills/nemotron-retrieval-recipes), [`nemotron-policy-generator`](skills/nemotron-policy-generator) |
| **Nemotron Speech** | Deploy and operate NVIDIA Nemotron Speech (Riva) NIMs — ASR, TTS, and NMT, cloud-hosted via build.nvidia.com or self-hosted on your own GPU. | [`nemotron-speech`](skills/nemotron-speech), [`nemotron-asr-finetune`](skills/nemotron-asr-finetune) |
| **Nemotron Voice Agent** | Create or refine NVIDIA voice agents (Cascaded or Omni) with Pipecat or LiveKit — speech customization, model selection, cloud or local deployment, and iteration. | [`nemotron-voice-agent-builder`](skills/nemotron-voice-agent-builder) |
| **Newton** | Agent skills for setting up, debugging, and optimizing Newton physics simulations using public documentation and packaged examples. | [`newton-usage`](skills/newton-usage), [`newton-performance`](skills/newton-performance) |
| **NIXL** | NVIDIA Inference Xfer Library (NIXL) — point-to-point data transfer for AI inference across GPU/CPU memory and storage: APIs, pluggable backends, builds, configuration, telemetry, troubleshooting, and benchmarks. | [`nixl-docs`](skills/nixl-docs) |
| **Nsight** | Agent-facing assets for Nsight developer workflows, including skills that help agents work with NVIDIA developer tools. | [`compute-sanitizer-diagnose-cuda-bugs`](skills/compute-sanitizer-diagnose-cuda-bugs), [`nsight-port-cpu-to-gpu`](skills/nsight-port-cpu-to-gpu) |
| **NVIDIA App** | NVIDIA App MCP: drivers, games, laptops, overlay. Check drivers, manage and optimize games, configure laptop features. | [`nvidia-app`](skills/nvidia-app) |
| **NVIDIA Broadcast** | Control NVIDIA Broadcast through its MCP gateway: apply audio and video effects, process local media, select devices, and change camera resolution. | [`nvidia-broadcast`](skills/nvidia-broadcast) |
| **NVIDIA FLARE** | Agent skills for converting training code to federated workflows, diagnosing jobs, collecting federated statistics, and running Auto-FL with NVIDIA FLARE. Install the complete NVIDIA FLARE skill set together so each workflow has its required shared references. | [`nvflare-autofl`](skills/nvflare-autofl), [`nvflare-autofl-report`](skills/nvflare-autofl-report), [`nvflare-convert-huggingface`](skills/nvflare-convert-huggingface), [`nvflare-convert-lightning`](skills/nvflare-convert-lightning), [`nvflare-convert-pytorch`](skills/nvflare-convert-pytorch), [`nvflare-diagnose-job`](skills/nvflare-diagnose-job), [`nvflare-fed-stats`](skills/nvflare-fed-stats), [`nvflare-orient`](skills/nvflare-orient), [`nvflare-shared`](skills/nvflare-shared) |
| **NVSHMEM** | Install, build, run, tune, and debug the NVSHMEM library and applications. | [`nvshmem-install`](skills/nvshmem-install), [`nvshmem-get-started`](skills/nvshmem-get-started), [`nvshmem-docs`](skills/nvshmem-docs), [`nvshmem-troubleshoot-and-report-bugs`](skills/nvshmem-troubleshoot-and-report-bugs), [`nvshmem-tune-performance`](skills/nvshmem-tune-performance), [`nvshmem-enable-tma`](skills/nvshmem-enable-tma), [`nvshmem-configure-nic-pe-mapping`](skills/nvshmem-configure-nic-pe-mapping), [`nvshmem-select-remote-transport`](skills/nvshmem-select-remote-transport), [`nvshmem-collect-performance-data`](skills/nvshmem-collect-performance-data) |
| **ovnewton** | Connect Newton physics to ovstage and Omniverse simulations. | [`ovnewton-usage`](skills/ovnewton-usage) |
| **Physical AI** | Physical AI skills for simulation, synthetic data generation, training, validation and deployment and more. | [`isaac-humanoid-cloud-control`](skills/isaac-humanoid-cloud-control), [`isaac-cable-insertion-policy-deployment`](skills/isaac-cable-insertion-policy-deployment), [`isaac-cable-insertion-policy-training`](skills/isaac-cable-insertion-policy-training), [`isaac-mission-control-showcase`](skills/isaac-mission-control-showcase), [`omniverse-cad-to-simready`](skills/omniverse-cad-to-simready), [`omniverse-content-services`](skills/omniverse-content-services), [`omniverse-realtime-viewer`](skills/omniverse-realtime-viewer), [`omniverse-usd-performance-tuning`](skills/omniverse-usd-performance-tuning), [`physical-ai-infrastructure-setup-and-resilient-scaling`](skills/physical-ai-infrastructure-setup-and-resilient-scaling), [`physical-ai-neural-reconstruction`](skills/physical-ai-neural-reconstruction), [`physical-ai-defect-image-generation`](skills/physical-ai-defect-image-generation), [`physical-ai-video-data-augmentation`](skills/physical-ai-video-data-augmentation), [`physical-ai-physics-selection`](skills/physical-ai-physics-selection) |
| **PhysicsNeMo** | NVIDIA PhysicsNeMo - Open-source deep-learning framework for building, training, and fine-tuning deep learning models using state-of-the-art Physics-ML methods. | [`physicsnemo-discover`](skills/physicsnemo-discover), [`physicsnemo-shard-tensor`](skills/physicsnemo-shard-tensor) |
| **Portfolio Optimization** | GPU-accelerated Mean-CVaR portfolio optimization with NVIDIA cuOpt — CVaR optimization, efficient frontier, scenario generation, backtesting, and rebalancing. | [`portfolio-optimization`](skills/portfolio-optimization) |
| **PVA SDK** | NVIDIA Programmable Vision Accelerator (PVA) SDK skills for environment setup, host readiness checks, operator specification, generation, build and test, and debugging. | [`pva-build-test`](skills/pva-build-test), [`pva-debug`](skills/pva-debug), [`pva-doctor`](skills/pva-doctor), [`pva-opgen`](skills/pva-opgen), [`pva-setup`](skills/pva-setup), [`pva-spec`](skills/pva-spec) |
| **RAG Blueprint** | RAG pipeline — deploy, configure, troubleshoot, and manage retrieval augmented generation with Docker Compose or Helm. | [`rag-blueprint`](skills/rag-blueprint), [`rag-eval`](skills/rag-eval), [`rag-perf`](skills/rag-perf) |
| **RTX Remix** | Mod and remaster classic games with RTX Remix: manage projects and USD layers, and replace textures and models through the Toolkit MCP server. | [`rtx-remix-modding`](skills/rtx-remix-modding) |
| **Run:ai** | Diagnose failed Run:ai workloads, summarize workload and cluster health, audit GPU quota fairness, and find idle GPU capacity through the Run:ai MCP server. | [`runai-diagnose-failed-workload`](skills/runai-diagnose-failed-workload), [`runai-get-workload-health-summary`](skills/runai-get-workload-health-summary), [`runai-get-utilization-overview`](skills/runai-get-utilization-overview), [`runai-audit-assigned-quota`](skills/runai-audit-assigned-quota), [`runai-get-cluster-health-summary`](skills/runai-get-cluster-health-summary) |
| **Skill Card Generator** | Reads an agent skill's source files and produces a skill card plus a review table. Use when a skill directory exists and a governance card needs to be generated or updated. | [`skill-card-generator`](skills/skill-card-generator) |

> The README is long; the above is its first 40000 characters. Full content: https://github.com/NVIDIA/skills
