BridgeAgent

An AI agent that treats the browser and the Windows desktop as one workspace. It reads web pages through the DOM and native windows through UI Automation, lets keyboard focus rather than the model decide which side to act on, replays learned skills without an LLM, and guards risky actions with permissions.把浏览器和 Windows 桌面当作同一个工作区的 AI 智能体。它通过 DOM 读取网页，通过 UI Automation 读取原生窗口，由键盘焦点而不是模型决定在哪一侧操作，学会的技能无需 LLM 即可回放，风险操作则由权限护栏把关。