智能体
x.com ·
@Andrew Ng (@AndrewYNg)
吴恩达：OpenWorker 基于 Nvidia OpenShell 为 AI Agent 加沙箱
吴恩达称 OpenAI-Hugging Face 被入侵源于沙箱隔离薄弱，并提到 Nvidia 开源了面向 AI Agent 的沙箱工具。他介绍其开源 Agent 框架 OpenWorker 将基于 Nvidia OpenShell，把每个 Agent 的命令执行限制在沙箱内：只放入任务相关文件，默认无法访问密钥、浏览器登录凭据和任意网站，限制由确定性代码而非 LLM 提示实现，并记录全部操作以供监控与审计。
AI推荐理由 吴恩达介绍 OpenWorker 基于 Nvidia OpenShell 为 AI Agent 提供沙箱隔离，并给出具体机制：仅挂载任务相关文件、默认屏蔽密钥与浏览器凭据、限制任意网站访问、用确定性代码而非提示词实现约束、全量日志审计。属于开发者可用的安全工程信息；但 OpenAI-Hugging Face 被黑一事仅为其个人表述，帖中未提供细节或证据，且 OpenWorker 的实际效果尚未验证。
80
浏览 117,155
点赞 1,568
评论 125
转发 216
正文
The OpenAI-Hugging Face hack was enabled by weak sandboxing. It is great that Nvidia is releasing open source tools for sandboxing AI agents. OpenWorker, our open-source agent harness supporting cybersecurity workflows, is proud to support this.
A sandbox gives an agent limited permissions. OpenWorker is building on Nvidia OpenShell and will support running each agent's commands inside a sandbox. Only the files relevant to the task go in. Secret API keys, your web browser login credentials, the ability to access arbitrary websites, are inaccessible to the agent by default. These restrictions are implemented in deterministic code rather than by prompting an LLM, which can make mistakes or be susceptible to prompt injections. Further, all actions are logged for monitoring and audit.
I'm grateful for @JensenHuang's leadership making AI agents more secure. OpenWorker (which @rohitcprasad and I are working on) will continue to improve security for agents.