Nvidia is giving OpenShell, a runtime for securing AI agents, to the Cloud Native Computing Foundation, and it published tools for running those agents on Kubernetes.
Nvidia said at KubeCon that it is giving OpenShell to the Cloud Native Computing Foundation as a sandbox project. OpenShell is an open source runtime for securing and governing autonomous agents. As a sandbox project, the foundation will host it, and the companies using it will help govern it.
Jim Zemlin, chief executive of the Linux Foundation, said agents need secure environments in which to operate. Nvidia said OpenShell applies set controls as agents use tools, reach inference services, and run workloads. It said hundreds of organizations have used it, and the project has drawn tens of thousands of contributions in less than a year.
Nvidia is also contributing to Agent Substrate, a Google-led framework already accepted into the same sandbox. The two companies are working on an integration so enterprises can secure agents across clouds and on their own systems.
On the same platform, Nvidia put Kubernetes Engine and Kubernetes Recipes in public preview, and released Snapshot, a tool that restores a saved inference replica without reloading the model. Nvidia said a Crusoe test of GLM-5.3 cut the time to ready from about eight minutes to 31 seconds. It said a Saturn Cloud test cut some startup times from between two and 8.5 minutes to between six and 34 seconds.
Tap a lens to see what this story means for you.
Liked this? The Brief brings you the whole day in tech, verified, every morning.
Two minutes, free forever. What's in The Brief →
See what’s happening right now
The Feed runs all day — short, verified briefs the moment they break.
Open the FeedFollow @thecircuitry_
Every story we publish, as it happens. No noise between.
Reader-supported
The Circuitry is a passion project I've always wanted to build, and I love the work behind it.
Running it costs real money. APIs, hosting, time. To keep improving the site and growing this into something useful for everyone, those costs have to be covered.
Any contribution is appreciated. If not, no pressure. Thanks for reading.
Nvidia's board says it is opening an independent investigation after reports that CEO Jensen Huang was the target of a foreign intelligence effort. It says it has no reason to believe secrets were compromised, and Huang has said he will cooperate.
Sources told CNBC that Chinese AI lab Moonshot has discussed leasing Nvidia Blackwell capacity from European neoclouds, including talks around clusters of 8,000 or more GPUs. No signed lease has been confirmed.
Linus Torvalds released Linux kernel 7.3 with a record 18,765 non-merge commits from 3,021 developers. The update adds features including sched_ext completion, failfs, and fchroot() while the 7.4 merge window opens next.
bbs-go through 4.4.6 has a Google One Tap login flaw (CVE-2026-111639) that skips ID-token audience checks, letting attackers replay tokens from any OAuth client to take over Google-linked accounts. VulnCheck rates it 8.1 (CVSS 3.1) and 8.6 (CVSS 4.0).
NVIDIA will celebrate the RTX Spark launch with global events on October 16, including in-person gatherings and an Instagram livestream. The systems combine up to 6,144-core Blackwell GPUs and 128 GB unified memory for local AI, creation, and gaming workloads on Windows PCs.
