做智能体开发和自动化工作流的团队终于有了专门的状态化执行环境——LangSmith Sandboxes 解决了智能体需要持久化工作状态和安全运行不可信代码的痛点,建议直接试用。
LangChain 发布了 LangSmith Sandboxes,专为 AI 智能体设计的执行环境,支持安装包、编辑文件、管理长期运行任务并恢复中断工作。该沙盒默认运行不可信代码,解决了智能体在复杂任务中需要持久化状态和安全隔离的核心痛点。开发者可以在沙盒中为智能体提供完整的计算环境,使其像人类一样持续工作。这一工具旨在提升智能体在自动化工作流、代码执行等场景中的可靠性和安全性。
Agents need stateful little computers where they can install packages, edit files, follow long-runni...
Agents need stateful little computers where they can install packages, edit files, follow long-running threads of work, and come back to where they left off. They need to run code that is untrusted by default. We built LangSmith Sandboxes specifically for this execution model. langchain.com/blog/langsmith… 💬 3 🔄 0 ❤️ 7 👀 923 📊 4 ⚡