NVIDIA launches OpenShell, an open secure runtime that governs AI agent execution outside the model
TheZachMueller · x · 2026-09-30
NVIDIA unveiled OpenShell, an open secure runtime for AI agents and the centerpiece of its Open Agent Safety Platform.
Key points:
- Security lives in the environment, not the model or app: deny-by-default, policy-based permissions enforced outside the agent's process, so controls can't be prompted away; every allow/deny is auditable.
- Model- and harness-agnostic: any model, any agent framework, governed through one unified policy layer.
- Flexible deployment: cloud, hybrid, on-premises, edge, and air-gapped environments.
An accompanying blog covers adding runtime controls to existing agents without rewriting their harness—protecting credentials, restricting access, and reviewing policy changes. The author plans to try it on a few idle Raspberry Pis for lightweight workloads.
More from coding & agent
- Generic API-catalog MCP server vs one tool per job: a context-bloat experiment — MandeepKaur9407 · 2026-11-16
- Berkeley paper: LLMs answer with stale values even when the update is in context — ombre-ai · 2026-11-16
- skills-gate: one gate skill fronts 148 agent skills at just 389 tokens — Distinct-Pie2389 · 2026-11-16
- Open-weight Karpo-Decision-31B tops Decision Index at 73.9, one forward pass per answer — CodeByPoonam · 2026-11-16
- shadcn boosts @uisdk: one-shot UI rebuild that's instantly E2E-testable by AI — shadcn · 2026-11-16
- AI agent deletes 43,599 lines of OpenClaw test code in one PR — vincent_koc · 2026-11-16