← Back to Skills Marketplace
wxd9199

Jetson CUDA

by kunka · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
158
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install jetson-cuda
Description
Provides NVIDIA Jetson CUDA tools to query CUDA version, GPU device info, and CUDA library paths on Jetson devices.
README (SKILL.md)

jetson-cuda

NVIDIA Jetson CUDA 工具集。提供 CUDA 版本查询、GPU 设备信息等功能。

功能

  • CUDA 版本查询 (nvcc --version)
  • GPU 设备信息 (nvidia-smi)
  • CUDA 库路径

使用方法

# 查看 CUDA 信息
jetson-cuda info

# 检查 nvcc
which nvcc
nvcc --version

依赖

  • CUDA Toolkit
  • nvidia-smi

环境要求

  • NVIDIA Jetson (AGX Orin, Xavier, Nano, etc.)
  • CUDA 11.x or 12.x
Usage Guidance
This skill appears to be what it says: a small helper that runs nvcc and nvidia-smi to report CUDA and GPU info. Before installing, be aware: (1) it will execute shell commands on the host (nvcc and nvidia-smi), so only enable it on devices you trust; (2) the metadata does not declare required binaries even though the script uses them—ensure your device actually has CUDA/nvcc or nvidia-smi if you expect useful output; (3) the skill performs no network communication and asks for no credentials. If you want extra caution, inspect scripts/cuda-info.sh yourself and run it manually in a terminal on a test device first.
Capability Analysis
Type: OpenClaw Skill Name: jetson-cuda Version: 1.0.0 The skill provides basic CUDA and GPU diagnostic information for NVIDIA Jetson devices. The script 'scripts/cuda-info.sh' only executes standard commands (nvcc and nvidia-smi) to retrieve version information, and the documentation in 'SKILL.md' is consistent with this functionality. No malicious or suspicious behaviors were identified.
Capability Assessment
Purpose & Capability
The skill name, description, SKILL.md, and the included script all align: they check nvcc and nvidia-smi to report CUDA/GPU info. Minor inconsistency: the registry metadata lists no required binaries, but the instructions and script call nvcc and nvidia-smi and the SKILL.md lists them as dependencies.
Instruction Scope
SKILL.md and scripts/cuda-info.sh only run local commands (which nvcc, nvcc --version, nvidia-smi) and print results or errors. There are no instructions to read unrelated files, transmit data externally, or access secrets.
Install Mechanism
This is instruction-only with a small included script and no install spec, so nothing is downloaded or installed by the skill itself.
Credentials
The skill requires no environment variables or credentials. That matches its functionality of querying local CUDA/GPU tools.
Persistence & Privilege
The skill does not request always: true and makes no changes to other skills or system-wide settings; it has ordinary, non-persistent scope.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install jetson-cuda
  3. After installation, invoke the skill by name or use /jetson-cuda
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of jetson-cuda. - 提供 CUDA 版本查询(nvcc --version) - 查询 GPU 设备信息(nvidia-smi) - 支持查看 CUDA 库路径 - 适用于 NVIDIA Jetson 平台(AGX Orin, Xavier, Nano 等) - 需要 CUDA Toolkit 和 nvidia-smi 依赖
Metadata
Slug jetson-cuda
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Jetson CUDA?

Provides NVIDIA Jetson CUDA tools to query CUDA version, GPU device info, and CUDA library paths on Jetson devices. It is an AI Agent Skill for Claude Code / OpenClaw, with 158 downloads so far.

How do I install Jetson CUDA?

Run "/install jetson-cuda" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Jetson CUDA free?

Yes, Jetson CUDA is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Jetson CUDA support?

Jetson CUDA is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Jetson CUDA?

It is built and maintained by kunka (@wxd9199); the current version is v1.0.0.

💬 Comments