[OpenCode Open Source Browser Control Tool: AI Takes Over Real Chrome, with Anti-Bot Detection Design] Jinse Finance reports that OpenCode has officially launched Browse, allowing its AI Agent to directly control browsers like Chrome that users use daily. It connects to tools like Codex, Claude Code, and OpenCode via MCP and browser extensions, reusing existing tabs, login states, and extensions without requiring a separate browser launch for the AI. The Agent can open its own tabs in the background, using Playwright to click, fill out forms, take screenshots, record screens, and log network requests without affecting normal user browsing. For steps requiring user intervention, such as login, two-factor authentication, or payments, control can be returned to the user. Browse also supports cross-device control. For example, the Agent can run on a remote server but control a user's logged-in Chrome account on their MacBook without migrating the password or the entire browser session to the cloud. Cross-device connections require additional encryption using tools like Tailscale. OpenCode developers have also added a dedicated anti-detection mechanism. The project uses Patchright to reduce the automated features exposed by the browser and simulates the rhythm of human mouse movement, typing, and scrolling by default, attempting to reduce the probability of being identified as a bot by websites. However, these measures cannot guarantee bypassing website risk controls. When encountering CAPTCHAs or anti-bot interceptions, Browse will identify anomalies and delegate the steps requiring manual verification to the user. The project is open source under the MIT license.