Dev & Engineering ✓ NVIDIA · Official holoscannvidia-sdkinstallationcudadockercondapython

Holoscan SDK Setup Guide

Inspects a Linux host and selects the most suitable Holoscan SDK installation path.

FollowSkills review · FSRS-2.0
Not recommended
55/ 100 5-point scale 2.8 / 5
Trust17 / 25 · 3.4/5

The skill uses read-only host inspection, explicitly forbids pasting installation commands before user confirmation, and delegates installation only after method selection; no malware, credential theft, or destructive default is evident. Points are deducted because external documentation, NGC, Docker, and GPU information are accessed without a detailed data-flow disclosure, least-privilege model, rollback plan, or granular confirmation for installation; dependency integrity verification is also unspecified.

Reliability8 / 20 · 2.0/5

The workflow, platform matrix, CUDA-variant rules, troubleshooting guidance, and delegation paths are mostly consistent, and the scripts provide some diagnostic output. Points are deducted because static review did not execute key paths; the scripts depend on the runtime and an unspecified run_script mechanism, the NGC check has thin error handling, and current platform/package rules still require external documentation confirmation.

Adaptability10 / 15 · 3.3/5

Audience, supported platforms, installation methods, limitations, and trigger scenarios are clearly described, including a negative evaluation task for an unrelated TensorRT question. Points are deducted because unsupported-platform boundaries remain partly vague, Chinese-language interaction is not specified, and the core flow depends on overseas documentation and NGC without discussing mainland-China reachability or alternatives.

Convention10 / 15 · 3.3/5

The skill includes versioning, name, license, author, dependencies, script documentation, limitations, troubleshooting, and evaluation records, with generally clear information architecture; official NVIDIA attribution and maintenance clues are present. Points are deducted for the missing recommended Examples section, incomplete changelog and update-path information, unclear long-term maintenance responsibility, and incomplete dependency version pinning; some installation details are delegated to other skills rather than verified here.

Effectiveness6 / 15 · 2.0/5

The skill can inspect the host, assess compatibility, compare methods, obtain user confirmation, and route installation to method-specific skills; the supplied evaluation report claims strong correctness and effectiveness. Points are deducted because those results are not independently verified in the source material, the skill itself does not perform installation, and usable completion depends on other skills, external documentation, Docker/NGC, and the host environment.

Verifiability4 / 10 · 2.0/5

The source includes explicit procedures, commands, scripts, evaluation tasks, and benchmark results, providing some auditability. Points are deducted because benchmark outcomes are report-level claims, no committed CI test suite covers this skill's key paths, and commands were not executed during this static review.

Evidence confidence:Low Reviewed Jul 20, 2026 Reviewed revision 55f18499943e
Before you use it
  • Confirm the reachability and trustworthiness of external documentation, NGC, Docker, and the NVIDIA Container Toolkit before installation, especially under mainland-China network constraints.
  • The skill reads host hardware, OS, driver, memory, tooling, and Conda information; review logging and agent-tool behavior to prevent disclosure of sensitive paths or environment details.
  • Do not treat the benchmark report as independently reproduced evidence; before execution, verify the current Holoscan release, CUDA compatibility, delegated-skill availability, and recovery options after failure.
  • Add representative examples, a changelog, dependency-version policy, recovery procedures, and sharper boundaries for unsupported platforms.
See the full review method →

What does this skill do, and when should you use it?

This skill helps users assess and choose an installation method for the NVIDIA Holoscan SDK. It inspects hardware, operating system, CUDA driver information, Docker, Python, apt, Conda, and existing Holoscan installations. It then compares NGC containers, Debian/apt, pip wheels, Conda, and source builds according to the detected platform and CUDA version. It is intended for Ubuntu, RHEL, IGX Orin, Jetson, and DGX Spark / Grace-Hopper installation decisions.

It first reads the current Holoscan SDK installation documentation to confirm supported platforms and requirements. It runs parallel checks for system identity, OS, architecture, nvidia-smi, CPU count, and memory, using the driver's reported CUDA Version to select variants. It checks Docker, Python, pip, apt, and existing Holoscan installations, verifies Docker GPU passthrough, and invokes check_conda.sh and check_ngc_image.sh for Conda and NGC image detection. It presents every supported installation method with host-specific status, recommends one method, asks the user to choose, and then delegates to the matching install skill.

  1. A first-time Ubuntu 22.04/24.04 x86_64 user wants the fastest Holoscan path on a workstation with Docker and an NVIDIA GPU.
  2. A user on RHEL 9.x needs to confirm why the NGC container is the only supported installation method.
  3. A Jetson or IGX Orin user needs a CUDA-variant-aware choice among container, apt, and source installation.
  4. A user cannot run conda --version but wants to find installed Conda environments that can import holoscan.
  5. An existing Holoscan user needs to identify the installed version and decide whether to upgrade or verify it.

What are this skill's strengths and limitations?

Pros
  • Covers container, apt, pip wheel, Conda, and source installation paths.
  • Uses the driver's CUDA Version to distinguish CUDA 12, CUDA 13, and ARM64 iGPU variants.
  • Verifies Docker GPU passthrough and includes dedicated scripts for Conda and NGC image detection.
  • States important platform constraints, including container-only RHEL 9.x and C++-only apt installs since Holoscan v3.0.0.
Limitations
  • Requires a Linux host, a working NVIDIA GPU driver, and network access to NVIDIA documentation and NGC.
  • The setup skill primarily assesses and delegates; the full installation command set belongs to separate Holoscan install skills.
  • Conda supports CUDA 13 only, while pip wheels require glibc ≥ 2.35 and Python 3.10–3.13.
  • The source provides no standalone test results or independent evidence beyond the listed platform guidance.

How do you install this skill?

Use the repository README's supported flow: npx skills add nvidia/skills --skill holoscan-setup --yes. You can also run npx skills add nvidia/skills --list to browse the catalog or npx skills update to refresh installed skills. The source does not document a separate manual-copy procedure for other clients.

How do you use this skill?

Install the skill folder where an Agent Skills-compatible client can load it, then ask: "Set up the Holoscan SDK on this Linux host, starting with platform compatibility and available installation methods." The skill inspects the environment, shows all options, asks for a method, and then delegates to holoscan-install-container, holoscan-install-debian, holoscan-install-wheel, holoscan-install-conda, or holoscan-install-source.

How does this skill compare with similar options?

Unlike a single fixed installation tutorial, this skill inspects the host first and conditionally recommends among NGC containers, apt, pip wheels, Conda, and source builds. The selected installation procedure is owned by the corresponding Holoscan install skill.

FAQ

Does this skill install Holoscan directly?
It inspects the environment and recommends a method first. After the user chooses, it delegates the actual installation to the matching install skill.
Why does it need the CUDA Version from nvidia-smi?
That field is the driver's maximum supported CUDA version and determines native package and container-tag selection. Native installations cannot exceed it.
Can Holoscan be installed without Docker?
On platforms that support native methods, apt, pip wheel, Conda, or source may be used. RHEL 9.x supports only the NGC container method.
Why might conda --version fail even when Conda is installed?
Conda may be installed in a common directory without being on PATH, or loaded lazily by shell initialization. check_conda.sh searches those directories and shell rc files.

More skills from this repository

All from NVIDIA/skills

Automation & Ops ✓ NVIDIA · Official

cuOpt Installation Guide

Install and verify NVIDIA cuOpt for Python, C, or REST deployment.

Automation & Ops ✓ NVIDIA · Official

Holoscan NGC Container Installer

Install, validate, and launch NVIDIA Holoscan SDK through an NGC Docker container.

Dev & Engineering ✓ NVIDIA · Official

Holoscan Python Wheel Installer

Install Holoscan SDK Python bindings in a virtual environment and verify them on a Linux NVIDIA GPU system.

Dev & Engineering ✓ NVIDIA · Official

Holoscan SDK Source Builder

Build Holoscan SDK from source in a Docker container and produce a local CMake-consumable installation tree.

Dev & Engineering ✓ NVIDIA · Official

Holoscan Native Ubuntu Installer

Install and verify the Holoscan SDK C++ environment on Ubuntu from NVIDIA’s apt repository.

Automation & Ops ✓ NVIDIA · Official

TAO NVIDIA GPU Host Setup

Checks and standardizes NVIDIA drivers, CUDA, and container runtime prerequisites for TAO GPU hosts.

Dev & Engineering ✓ NVIDIA · Official

Earth2Studio Deterministic Forecast

Build single-member weather forecast inference scripts with Earth2Studio.

Dev & Engineering ✓ NVIDIA · Official

CuTile Autotuning Guide

Design, implement, and validate low-overhead autotuning for CuTile GPU kernels.

Data & Analysis ✓ NVIDIA · Official

TAO AOI Image Mining

Embed target and source images with one encoder, then mine deduplicated nearest-neighbour AOI images for augmentation.

Dev & Engineering ✓ NVIDIA · Official

Holoscan Conda Installer

Install and validate Holoscan SDK on Linux x86_64 with a CUDA 13 driver using Conda.

Dev & Engineering ✓ NVIDIA · Official

NeMo Relay Installation Guide

Choose and verify the right NeMo Relay path for CLIs, language packages, and maintained frameworks.

Dev & Engineering ✓ NVIDIA · Official

DALI Dynamic Mode Assistant

Helps agents write, review, and migrate NVIDIA DALI imperative dynamic-mode code.

Dev & Engineering ✓ NVIDIA · Official

cuPyNumeric Installation Guide

Guides safe cuPyNumeric installation and verifies that it actually works.

Automation & Ops ✓ NVIDIA · Official

Jetson Package & Environment Advisor

Select Jetson-compatible containers and GPU package indexes by device architecture.

Data & Analysis ✓ NVIDIA · Official

TAO Grounding DINO Open-Vocabulary Detection

Train and deploy text-guided object detectors without a fixed class vocabulary.

Dev & Engineering ✓ NVIDIA · Official

CUDA-Q Quantum Onboarding

Guides developers from CUDA-Q installation to quantum kernels, GPU simulation, and real QPU execution.

Automation & Ops ✓ NVIDIA · Official

Nemotron Speech NIM Skill

Guides deployment and operation of NVIDIA Riva ASR, TTS, and translation NIMs.

Dev & Engineering ✓ NVIDIA · Official

NVIDIA Skill Finder

Finds the right NVIDIA agent skill for product, hardware, and workflow requests.

Dev & Engineering ✓ NVIDIA · Official

TAO Sparse4D Training Skill

Guides agents through training, evaluation, export, quantization, and inference for multi-camera temporal 3D detection and tracking.

Dev & Engineering ✓ NVIDIA · Official

i4h Catheter Navigation Workflow Guide

Orients agents to the right stage of NVIDIA’s endovascular catheter navigation workflow.

Related skills