{"slug": "nvidia-openshell", "title": "Nvidia OpenShell", "summary": "NVIDIA released OpenShell 0.1.x, an open-source runtime for executing fleets of autonomous AI agents in sandboxed environments with kernel-level isolation. The release adds a stable release cadence, new isolation primitives, and an expanded extension surface, and pairs sandbox runtime controls with a declarative YAML policy so teams can run agents without granting unrestricted access to local files, credentials, and external networks. An upgrade guide documents everything that changed in the 0.1.0 release.", "body_md": "# Overview of NVIDIA OpenShell\n\nNVIDIA OpenShell is an open-source runtime for executing fleets of autonomous AI agents in sandboxed environments with kernel-level isolation. It combines sandbox runtime controls and a declarative YAML policy so teams can run agents without giving them unrestricted access to local files, credentials, and external networks.\n\nNew in OpenShell 0.1.x: a [stable release cadence](https://docs.nvidia.com/openshell/about/support-matrix#releases), new [isolation primitives](https://docs.nvidia.com/openshell/about/architecture), and an expanded [extension surface](https://docs.nvidia.com/openshell/extensibility/overview), along with much more.\n\nSee our [upgrade guide](https://docs.nvidia.com/openshell/upgrade/0-1-0) for everything that’s changed.\n\n## Why OpenShell Exists\n\nAI agents are most useful when they can read files, install packages, call APIs, and use credentials. That same access can create material risk. OpenShell is designed for this tradeoff: preserve agent capability while enforcing explicit controls over what the agent can access.\n\n## Common Risks and Controls\n\nThe table below summarizes common failure modes and how OpenShell mitigates them.\n\n## Protection Layers at a Glance\n\nOpenShell applies defense in depth across the following policy domains.\n\nFor details, refer to [Sandbox Policies](https://docs.nvidia.com/openshell/how-it-works/policies/overview) and [Default Policy](https://docs.nvidia.com/openshell/how-it-works/policies/default-policy).\n\n## Common Use Cases\n\nOpenShell supports a range of agent deployment patterns.\n\n## Next Steps\n\nExplore these topics to go deeper:\n\n- To understand the runtime architecture, refer to [Architecture](https://docs.nvidia.com/openshell/about/architecture) .\n- To prepare an image and launch an agent, refer to [Run Your First Agent](https://docs.nvidia.com/openshell/about/run-your-first-agent) .\n- To learn how OpenShell enforces policy controls across protection layers, refer to [Sandbox Policies](https://docs.nvidia.com/openshell/how-it-works/policies/overview) .", "url": "https://wpnews.pro/news/nvidia-openshell", "canonical_source": "https://docs.nvidia.com/openshell/about/overview", "published_at": "2026-10-01 11:17:13+00:00", "updated_at": "2026-10-01 11:46:07.803521+00:00", "lang": "en", "topics": ["ai-agents", "ai-safety", "ai-tools", "developer-tools", "artificial-intelligence"], "entities": ["NVIDIA", "NVIDIA OpenShell"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/nvidia-openshell", "markdown": "https://wpnews.pro/news/nvidia-openshell.md", "text": "https://wpnews.pro/news/nvidia-openshell.txt", "jsonld": "https://wpnews.pro/news/nvidia-openshell.jsonld"}}