Browser Control
Local browser automation lets coding agents use Playwright with your existing Chromium profile, logged-in sessions and extensions. Open source under MIT.
Browser Control gives trusted coding agents access to the Chromium browser you already use, with its signed-in accounts and installed extensions. It's for developers who want agents to work in an existing browser session rather than a separate headless browser. The driver runs locally through a browser extension and local relay. It contains no LLM.
Agents can connect through a CLI or MCP, including clients such as OpenCode and Claude Code. Browser support includes Chrome, Brave, Edge, Arc and Chromium; the local runtime requires Node.js. Persistent sessions retain a page and JavaScript state across calls, while agents can also take control of an existing tab. Releasing that tab leaves it open.
Page snapshots, annotated screenshots and visual comparisons help agents inspect a page before acting and check the result afterward. Element references reject changed targets rather than silently selecting a different control. Human handoffs pause automation for CAPTCHA, two-factor authentication or payment confirmation. Demonstrations can turn a workflow performed by a person into editable Playwright code.
For developers building applications, the TypeScript client can make authenticated requests using the browser's cookies without moving those cookies out of the browser. Network capture can also produce direct API clients, with credentials stored separately from generated code and redacted in exported artifacts. Activity journals stay local. The project is open source under the MIT license; its read-only mode blocks mouse and keyboard input but isn't a security sandbox for untrusted code.
