keon/browser-control

What is it?

Key features

• 01Snapshot returns stable element refs that can be used directly in click, fill, and other actions.
• 02Every capability is a shell subcommand printing compact text or JSON.
• 03Raw CDP access through eval and cdp commands covers operations not wrapped by helpers.
• 04A background daemon keeps recent events, network requests, and console messages.
• 05Failed commands write diagnostic traces under the workspace directory.
• 06Supports local Chrome and cloud browser providers with the same commands.

Use cases

• Letting a coding agent inspect and interact with a web page from its shell.
• Scripting repeatable browser steps such as logins or form filling in a shell pipeline.
• Running a browser session on a cloud provider like Browser Use or Steel from the command line.
• Debugging a browser automation flow by reviewing the saved failure traces.

Quick start

1. 1

   Install the CLI from crates.io

   cargo install browser-control-cli
2. 2

   Create the workspace and launch Chrome on an example page

   browser-control init && browser-control launch https://example.com
3. 3

   Take a snapshot of the page to list actionable element refs

   browser-control snapshot

Things to know

• !Needs a local Chrome or Chromium build, or a paid cloud provider API key for remote sessions.
• !The crate is named browser-control-cli because the shorter name was taken.

In collections

Topics

FAQ

What is browser-control?

A small Rust CLI that drives a real Chrome browser over the Chrome DevTools Protocol, designed for shell-based coding agents. browser-control is a command-line tool written in Rust that exposes a browser as a set of shell subcommands. It can launch or connect to a local Chrome instance, take page snapshots that assign stable element refs such as @e1 and @e2, and then click, fill, type, scroll, or upload against those refs. CSS selectors and x,y coordinates are also accepted as targets.

Is browser-control free to use? What license does it have?

browser-control is open source under the MIT license. Check the license terms before using it in commercial products.

Is browser-control still maintained?

Updated within 6 months. The last commit to browser-control was on Aug 27, 2026.

What can I use browser-control for?

Letting a coding agent inspect and interact with a web page from its shell. Scripting repeatable browser steps such as logins or form filling in a shell pipeline. Running a browser session on a cloud provider like Browser Use or Steel from the command line. Debugging a browser automation flow by reviewing the saved failure traces.

How do I install browser-control?

Install the CLI from crates.io: cargo install browser-control-cli Create the workspace and launch Chrome on an example page: browser-control init && browser-control launch https://example.com Take a snapshot of the page to list actionable element refs: browser-control snapshot

Where can I get browser-control?

The source code and installation instructions are on GitHub: https://github.com/keon/browser-control