skills.rest
NVIDIA
NVIDIAOFFICIAL
Explore all skills by this author →
WebsiteGitHub

jetson-video-setup

Installs, probes, and verifies NVIDIA Video Codec SDK and PyNvVideoCodec on Jetson devices.

3.2k|370|Updated Feb 25, 2026
Search Tags:#h264#setup verification#jetson#NVENC#nvdec#pynvvideocodec#video codec sdk
One-click install
npx skills add https://github.com/NVIDIA/skills --skill jetson-video-setup
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: jetson-video-setup
Source: https://github.com/NVIDIA/skills/tree/main/skills/jetson-video-setup
Command: npx skills add https://github.com/NVIDIA/skills --skill jetson-video-setup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires PyNvVideoCodec, pycuda, numpy, torch, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Setting up hardware video encoding and decoding on NVIDIA Jetson devices involves fragile steps: selecting the right surface (native Video Codec SDK vs. PyNvVideoCodec), authenticating APT package sources, creating isolated Python environments, and proving the installation actually works. This Skill automates that probe → plan → install → verify workflow with official encode-to-decode smoke tests.

Core Features & Use Cases

  • Read-only environment probing: Captures Jetson Linux release, GPU, APT candidates, CUDA toolkit, and Python state into a signed JSON artifact before any change.
  • Authenticated installation planning: Generates and validates install plans bound to signature-verified NVIDIA Jetson APT sources, with hash-locked pip installs for PyNvVideoCodec in a fresh venv.
  • Official operation proof: Verifies readiness by encoding one 640×360 NV12 frame to H.264 with the installed release's official samples and independently decoding the fresh bitstream.
  • Use Case: Ask your agent to install PyNvVideoCodec in a new isolated environment on a Jetson and run an H.264 encode/decode smoke test; the Skill provisions the venv, applies the locked pip install, and reports a ready verdict only after the official samples pass.

Quick Start

Ask your agent to install the native Video Codec SDK on this Jetson and verify it with an H.264 encode and decode smoke test.

Frequently Asked Questions about jetson-video-setup

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I install PyNvVideoCodec on a Jetson device?▼

PyNvVideoCodec 2.1.0 is installed into a newly created isolated virtual environment using a hash-locked pip apply, after probing the target and validating an install plan. The setup then runs the wheel-owned encode and decode samples to prove the environment works before registering it.

How do I verify the NVIDIA Video Codec SDK installation on Jetson?▼

Verification builds the package-owned AppEncCuda and AppDec samples, encodes one 640×360 NV12 frame to H.264, and independently decodes the fresh bitstream. Readiness requires the exact positive markers and a 345,600-byte decoded NV12 output, not just a zero exit code.

What is the difference between Video Codec SDK and PyNvVideoCodec setup?▼

The native Video Codec SDK is installed as the nvidia-video-codec-sdk APT package from the official Jetson repository, while PyNvVideoCodec is a Python wheel installed into an isolated venv. They are independent products, and a request for one never installs the other.

Can I reuse an existing PyNvVideoCodec environment instead of reinstalling?▼

Yes, the registered validated environment is reused without any reinstallation when it is still ready. If the registry is absent or stale, you must supply the exact interpreter path; the workflow never scans the filesystem or guesses among virtual environments.

Why does the setup refuse to install packages from my APT source?▼

Native SDK and CUDA packages are accepted only from the signature-authenticated official NVIDIA Jetson sources at repo.download.nvidia.com with the exact rNN.N/main suite. Internal mirrors, unsigned candidates, HTTP sources, or trust bypasses are rejected, and sources or keys are never modified.

Does this setup skill benchmark codec performance or check codec support?▼

No, setup only proves installation readiness through official encode-to-decode smoke tests. Codec capability matrices, recipes, benchmarks, and pipelines are delegated to the sibling jetson-video-capability, recipe, benchmark, and pipeline skills.

Related Skills

Explore More Software Engineering Skills→
NVIDIA-AI-IOTNVIDIA-AI-IOT
46

jetson-download-bsp

Automate downloading NVIDIA Jetson Linux BSP artifacts for the active target platform.

Official
Intermediate
dongg622dongg622

mxffmpeg

Transcode videos between H.264, H.265, and AV1 with GPU acceleration.

Community
Intermediate
ldk-hubldk-hub

bmad-prfaq

Generate a PRFAQ document with press release, FAQs, verdict, and PRD distillate.

Community
Advanced
JORDAN0615JORDAN0615

jetson-download-bsp

Download NVIDIA Jetson Linux BSP artifacts for the active target platform to the correct workspace paths.

Community
Intermediate