NVIDIA agent skills
Develop GPU computing, AI, simulation, and robotics workflows with NVIDIA-maintained and verified skills.
Published by NVIDIA
What these skills help with
Develop GPU computing, AI, simulation, and robotics workflows with NVIDIA-maintained and verified skills.
Work with GPU data
Use accelerated computing tools for data processing.
Build AI infrastructure
Follow setup and scaling guidance for AI workloads.
Configure robotics workflows
Prepare simulation and camera calibration workflows.
Explore NVIDIA skills
View all
Accelerated Computing 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.
Amc Run RTSP Calibration
Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API.
Amc Run Sample Calibration
Run end-to-end calibration on the shipped sample dataset (sdg082sampledata010926.zip) against a running AMC microservice.
Install NVIDIA skills
Download the complete source, choose a skill folder, and follow the setup instructions for your assistant.
Read the source repository
Check prerequisites, supported tools, and installation steps from the publisher.
nvidia/skillsDownload the source files
Keep scripts and references with the SKILL.md file. Some skills require a service account or API credentials.
Download source ZIP
NVIDIA skill workflows
Explore a quantum circuit
Follow CUDA-Q guidance to set up and run a quantum computing example.
Solve an optimization problem
Use cuOpt guidance to define objectives and review solver results.
Prepare a Jetson device
Follow the quick-start instructions for an edge AI development setup.
Agent skills FAQ
An agent skill is a folder of instructions, usually headed by a SKILL.md file, that helps an AI assistant perform a specific task. A skill may also include scripts, examples, and reference files.
