zenbu-labs/terminal-browser

What it solves

terminal-browser brings a full-featured graphical web browser directly into your terminal. It is designed to bridge the gap between terminal-based development environments and the web, allowing developers to interact with websites, debug applications, and provide web-based context to AI coding agents without leaving their command-line interface.

How it works

The project uses the Kitty graphics protocol to render pixels generated by a headless Chromium instance directly inside the terminal window. It captures user input—such as mouse clicks, scrolling, and keyboard events—and translates them into synthetic events for the browser engine. The UI is built using a custom Rust-based engine and React, allowing for seamless layering of browser content and terminal-based UI elements. It also includes an API for terminals to implement, enabling advanced features like splitting panes, managing tabs, and direct integration with coding agents.

Who it’s for

This tool is intended for developers and power users who spend their time in terminal-based workflows (like Neovim, Tmux, or VS Code) and want to integrate web browsing or AI-driven web interaction into their existing terminal layout.

Highlights

• Agent-Ready: Designed to work with coding agents, allowing them to interact with the browser, view page content, and perform web-based tasks.
• High Performance: Uses offscreen rendering and sends only changed pixel patches to the terminal to ensure smooth, frame-rate-stable performance.
• SSH Support: Includes a proxy mode that runs the browser locally while routing network requests through a remote machine, enabling easy previewing of remote web applications.
• Extensible API: Provides a socket-based API for terminal emulators to implement, allowing for deep integration with terminal multiplexers and window managers.