NVIDIA just dropped a big batch of open-source “physical AI” skills and tools, and they’re designed to make a roboticist’s life a whole lot easier. The idea? Take the messy, complicated work behind robots, self-driving cars, vision AI, and industrial digital twins, and break it into bite-sized tasks that AI agents can actually run themselves.
These skills ship as part of the NVIDIA Agent Toolkit, and here’s what makes them handy: they let AI agents tap directly into NVIDIA’s own libraries, models, and frameworks. That means agents can help speed up the whole pipeline, from generating data and running simulations to training models, evaluating results, and finally deploying everything that powers robots, autonomous vehicles, factories, and labs.
“AI agents are revolutionizing software development, and that shift is now coming to physical AI, extending into the systems that will transform transportation, manufacturing, healthcare and robotics,” said Jensen Huang, founder and CEO of NVIDIA. “When agents can directly use NVIDIA libraries, models and frameworks, physical AI development will move faster, enabling developers to build the robots, autonomous vehicles and industrial systems of the future at an incredible pace.”
NVIDIA is reworking its entire physical AI stack with agents in mind, basically turning its libraries, models, and frameworks into tools that agents can call on directly. That covers a lot of ground: Cosmos world foundation models for reasoning about and generating the physical world, Omniverse libraries for simulation and digital twins, Isaac for robotics simulation and robot learning, Metropolis for vision AI, Alpamayo for autonomous driving, and the Jetson platform for building edge AI.
To help developers actually put all of this to work, NVIDIA is rolling out new skills inside the NVIDIA Agent Toolkit. The goal is to turn physical AI development into repeatable, step-by-step instructions that coding agents can follow, spelling out which tools to call, what outputs to generate, and how developers can check that the results hold up.
There’s a safety angle too. Using these skills, developers can build and deploy autonomous agents through the NVIDIA NemoClaw blueprint and the NVIDIA OpenShell runtime, which add policy-based security and privacy governance whether you’re running on local or cloud hardware.
NVIDIA physical AI skills and tools are accelerating agentic development across:
Robotics and edge AI: Robot developers can use skills to accelerate the entire robotics development pipeline, from generating perception and mobility training data to simulation, automating navigation training, advancing robot learning and tuning Jetson-based edge systems for deployment.
Autonomous vehicles: For AV developers, skills can direct agents to reconstruct data captured by fleets into simulation environments, generate photorealistic driving scenarios at scale and run closed-loop reinforcement learning to expand training and evaluation coverage.
Real-time vision AI agents: For automated inspection and video intelligence, agent skills help teams generate synthetic training data, fine-tune models, automate labeling and build video AI agents that search, summarize and analyze live or recorded video.
Industrial AI: Industrial software developers can use these skills to convert engineering data into computer-aided design (CAD) assets for digital twin simulation, optimizing large OpenUSD scenes with less manual setup.
Healthcare: Before deploying automation in clinical environments, healthcare teams can guide agents through hospital-environment digital twin creation, sim-to-real data generation and software-in-the-loop policy testing.
The skills can be combined and integrated into larger agentic systems, enabling developers to orchestrate and automate complex workflows such as data generation, simulation, optimization, inference tuning, continuous evaluation and more.
NVIDIA physical AI agent tools and skills are now openly available through GitHub and skills.sh for use with any coding agent.
Download: Secure Foundations for AI Workloads on AWS
