Tagged articles

OpenShell

2 articles · Page 1 of 1
Geek Labs
Geek Labs
Oct 27, 2026 · Artificial Intelligence

NVIDIA OpenShell: Kernel-Level AI Agent Sandbox with SMT-Verified Policies

NVIDIA's OpenShell secures AI agents by combining kernel-enforced sandbox isolation with SMT-solver-based formal verification of access policies, letting agents request permissions that are mathematically checked before human approval, while supporting Docker, Kubernetes, and MicroVM runtimes.

AI agent securityDockerKubernetes
0 likes · 22 min read
NVIDIA OpenShell: Kernel-Level AI Agent Sandbox with SMT-Verified Policies
AI Waka
AI Waka
Mar 25, 2026 · Information Security

How NemoClaw Secures Autonomous AI Agents with Kernel‑Level Sandboxing

This article examines NemoClaw’s three‑layer architecture that adds kernel‑level sandboxing, policy‑driven deployment, and flexible inference routing to OpenClaw, outlines installation steps, compares it with the native OpenClaw runtime, and discusses current limitations for production use.

AI agent securityInfrastructureNemoClaw
0 likes · 9 min read
How NemoClaw Secures Autonomous AI Agents with Kernel‑Level Sandboxing