NVIDIA/skills is a catalogue of portable instruction sets that teach coding agents how to use NVIDIA software for robotics, simulation, CUDA, retrieval-augmented generation, and related workflows. Developers install these skills in agents such as Claude Code or Codex, while the catalogue mirrors skills maintained in separate NVIDIA product repositories.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add NVIDIA/skills --skill hsb-appgit clone --depth 1 https://github.com/NVIDIA/skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nvidia/skills/hsb-app)<a href="https://agentmods.dev/skills/nvidia/skills/hsb-app"><img src="https://agentmods.dev/badge/skills/nvidia/skills/hsb-app.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00056 | $0.04075 |
| Opus 5 | $0.00028 | $0.02037 |
| Sonnet 5 | $0.00011 | $0.00815 |
| Haiku 4.5 | $0.00006 | $0.00407 |
Grade D, and why
hsb-app scanned grade D with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
REMOTE_SUDO sudo / sudo -n / "" — default to "sudo" if not set. Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
ssh -o BatchMode=yes $REMOTE_SSH_OPTS $SSH_TARGET "rm -rf /tmp/.claude_hsb_app_session" How it starts
The opening of the file, as written. The whole thing — 396 lines — stays where its author put it; the contents beside it link to each section on GitHub.
HSB Application Runner
Use this skill when the user wants to discover, select, and run Holoscan Sensor Bridge example applications on a devkit with a connected HSB board.
This skill assumes the devkit is already set up (SSH, demo container built, host configured, board connected). If setup is not complete, instruct the user to run /hsb-setup first.
This workflow runs applications inside the demo container. Only run it when the user explicitly invokes it.
Before you start — required gates (do these first, in order)
Gate 1 — Read environment variables. Before doing anything else, check these variables and print their resolved values to the user:
SSH_TARGET Remote devkit login (e.g. [email protected]). Ask the user if not set.
REMOTE_ROOT Remote working directory (e.g. /home/nvidia). Ask the user if not set.
REMOTE_SUDO sudo / sudo -n / "" — default to "sudo" if not set.
REMOTE_SSH_OPTS Additional SSH options (optional).
HSB_PLATFORM Platform hint (optional).
SSH_TARGET and REMOTE_ROOT are required. Stop and ask the user for them if either is missing.
Gate 2 — Present the phase plan and get confirmation. Before taking any action:
If the user's request already includes platform, board type, and sensors, also state upfront:
- You will scan
examples/and filter apps by the user's sensor type and platform - You will NOT add
--headlessautomatically — only if the user explicitly requests it - If the user specified a timeout (e.g., "60-second timeout"), state you will use that as the watchdog timeout
- Applications run inside the demo container via
docker run, usingpython3for Python-based examples
Show the phase plan:
HSB App — Phase Plan
Phase 0: Verify board connectivity and demo container readiness
Phase 1: Discover user setup and select application to run
Phase 2: Run application with monitoring, failure analysis, and iterative debugging
Phase 3: Generate session report (with option to save)
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 396 lines · 56 tokens per session scan D 4fc440a72053
hsb-app is a skill published in the GitHub repository NVIDIA/skills (3,211 stars, last pushed today), licensed Apache-2.0. It adds 56 tokens to every session and 4,075 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it D with 2 findings (asks for root, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
terraform-module-library
Build reusable Terraform modules for AWS, Azure, GCP, and OCI infrastructure following infrastructure-as-code best practices. Use when creating infrastructure modules, standardizing cloud provisioning, or implementing reusable IaC components.
hyperpod-version-checker
Check and compare software component versions on SageMaker HyperPod cluster nodes - NVIDIA drivers, CUDA toolkit, cuDNN, NCCL, EFA, AWS OFI NCCL, GDRCopy, MPI, Neuron SDK (Trainium/Inferentia), Python, and PyTorch. Use when checking component versions, verifying CUDA/driver compatibility, detecting version mismatches…
technical-troubleshooting
Provide setup, troubleshooting, and maintenance guidance. Use when the user reports a device that won't power on, connectivity issues, setup questions, overheating, or maintenance concerns.
gpu-optimizer
GPU optimization for consumer NVIDIA GPUs (8-24GB VRAM) covering mixed precision, gradient checkpointing, XGBoost GPU, CuPy/cuDF migration, and torch.compile. Triggers on: "optimize GPU training", "speed up CUDA", "reduce OOM", "migrate NumPy to CuPy", "manage GPU memory", "benchmark PyTorch".