做 AI 智能体开发的团队终于可以省去环境搭建的麻烦——单次 API 调用就能跑起一个完整智能体，建议直接试试。
Google 在 Gemini API 中推出了 Managed Agents 功能，旨在简化 AI 智能体的构建。开发者只需一次 API 调用，即可创建一个能在托管 Linux 沙箱中推理、编写和运行代码、管理文件的智能体。该服务将循环和环境管理移至平台端，让开发者专注于产品构建而非基础设施。团队表示将快速迭代并欢迎反馈。
Together with @OfficialLoganK and @alihcevik we sat down and discussed the launch of Managed Agents ...
Together with @OfficialLoganK and @alihcevik we sat down and discussed the launch of Managed Agents in the Gemini API. We wanted to make building AI Agents much simpler. With just a single API call, you can now spin up an agent that reasons, writes and runs code, and manages files inside a hosted Linux sandbox. Moving the loop and environment into the platform, let you focus on building your product instead of managing infrastructure. Tag me with any feedback, questions or ideas. We want to iterate fast and make it better every day. 👇 Your browser does not support the video tag. 🔗 View on Twitter 💬 2 🔄 2 ❤️ 10 👀 540 📊 4 ⚡