Nvidia said on November 10, 2026, at KubeCon + CloudNativeCon, that it is contributing OpenShell to the Cloud Native Computing Foundation (CNCF), part of the Linux Foundation. OpenShell is an open-source runtime for securing and governing autonomous agents. As a CNCF Sandbox project, it will be governed by the organizations building it into their products rather than by Nvidia alone.
OpenShell applies deterministic controls to what agents do, so teams can enforce policies as agents call tools, reach inference services and run workloads. Nvidia says hundreds of organizations have used it and that it has received tens of thousands of contributions in less than a year since launch. Nvidia and Google are also working on an integration between OpenShell and Agent Substrate, a Google-led framework for running long-lived agent workloads on Kubernetes. Agent Substrate has been accepted into the CNCF Sandbox, and Nvidia is contributing to it.
Nvidia also announced NVIDIA Kubernetes Engine (NVKE) and NVIDIA Kubernetes Recipes (NVKR), building blocks for managed Kubernetes services on GPU infrastructure, now in public preview. They automate cluster lifecycle management and provide signed, version-pinned images. Ecosystem partners include Mirantis, Netris, OpenNebula Systems, Rafay, Spectro Cloud, vCluster, Firmus and SK Telecom.
The third release is NVIDIA Snapshot, an open-source component that restores checkpointed inference replicas in seconds without reloading or warming up the model. Nvidia says Crusoe, testing GLM-5.3 on HGX B300 systems, cut time-to-ready from about eight minutes to 31 seconds. Saturn Cloud cut vLLM startup times from between two and 8.5 minutes to between six and 34 seconds.
For teams putting agents into production, moving OpenShell to a neutral foundation makes it easier to adopt as a shared policy layer without depending on a single vendor. Snapshot targets a separate problem: how long it takes to scale inference capacity up when demand spikes.