Nvidia 发布 Open Agent Safety Platform，用硬件隔离监控失控智能体
Nvidia 把智能体的“急停开关”做进了硬件，OpenShell 开源，Sentry 跑在 BlueField-4 上，毫秒级就能掐住失控的 agent，思路挺狠。
Nvidia 推出 Open Agent Safety Platform，把智能体监控从软件层移到智能体无法触及的硬件层。软件部分 OpenShell 是 Apache 2.0 开源运行时，负责沙箱隔离并对文件、网络、工具和凭据执行操作者设定的限制。硬件部分 NVIDIA Sentry 是基于 BlueField-4 DPU 的参考设计，能在毫秒级隔离失控智能体。在 Vera Rubin POD 中，每个机柜的 BlueField-4 位于节点访问模型的唯一通路上，与主机隔离，因此能监控智能体的每一步并直接切断。
Nvidia dropped an open safety platform built to catch and quarantine rogue agents in real time before they escape.
NVIDIA's new Open Agent Safety Platform moves AI agent monitoring onto a chip the agent can't reach.
Its software half, OpenShell, is an Apache 2.0 runtime that sandboxes agents and enforces operator limits on files, networks, tools and credentials.
Its hardware half, NVIDIA Sentry, is a reference design running on BlueField-4 data processing units that can quarantine straying agents within milliseconds.
In a Vera Rubin POD, each tray's BlueField-4 sits on the node's only route to the model, isolated from the host. Because agents must consult the model before acting, that position gives Sentry a view of every step and a kill switch.
