Nvidia Launches Open-Source Toolkit for Physical AI Agents on GitHub
Nvidia introduced an open-source collection of physical AI agent skills and tools at its GTC event, making the package available on GitHub for agentic and physical AI development. The collection is designed to let agents tap Nvidia technologies including Omniverse libraries, Cosmos world foundation models, Isaac simulation frameworks, Metropolis and Jetson, with use cases spanning autonomous vehicles, robotics, vision AI, industrial digital twins and healthcare.
The rollout also includes security tooling for the software that agents use. Nvidia's SkillSpector scanner is designed to check AI agent skills before installation, running 64 security checks across 16 categories, including prompt-injection, credential-theft and supply-chain vulnerability detection, with static analysis, optional LLM evaluation and SARIF output for CI/CD.
From the sources (7 posts)
@nptacekRT @bibryam: SkillSpector - a new security scanner for skills by NVIDIA • Scan AI agent skills before installing them • 64 security checks…
@bibryamSkillSpector - a new security scanner for skills by NVIDIA • Scan AI agent skills before installing them • 64 security checks across 16 categories • Fast static analysis + • Optional LLM semantic evaluation • Prompt injection detection • C
@nvidianewsroomAt #NVIDIAGTC, Jensen announced a major open source collection of physical AI agent skills and tools, available now on GitHub. The new collection helps agents tap into NVIDIA technologies including Omniverse libraries, Cosmos world foundat
@financialjuiceNvidia unveils extensive open source toolkit for physical AI agents
@firstsquawkNVIDIA has launched a large set of open-source tools and capabilities for agentic and physical AI development.
@firstsquawkNVIDIA introduces a new AI agent toolkit built around Nemoclaw, Nemotron, OpenShell, and CUDA-X.
@financialjuiceNvidia launches agent toolkit with Nemoclaw, Nemotron, Openshell, and Cudax