borz

Control a Chromium browser via Chrome DevTools Protocol from a single binary.

1|Updated Apr 15, 2026
One-click install
npx skills add https://github.com/leolin310148/borz --skill borz
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: borz
Source: https://github.com/leolin310148/borz/tree/main
Command: npx skills add https://github.com/leolin310148/borz --skill borz

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

borz solves the challenge of programmatically automating and observing a live Chromium-based browser, including authenticated sessions, JavaScript rendering, and multi-step interactions.

Core Features & Use Cases

  • Three front-ends (CLI, MCP, REST) share one persistent browser session to enable automation, AI control, and integration with workflows.
  • Snapshot-based interaction: obtain element refs via the accessibility tree, then drive actions like open, click, fill, type, and eval against a real browser.
  • Site adapters and extension-backed APIs unlock cross-domain cookies, bookmarks, and browser state for end-to-end automation.

Quick Start

Launch borz and start a daemon, then open a URL to begin interactive automation with a live Chrome session.

Frequently Asked Questions about borz

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate a live Chrome browser session from the command line?

To automate a live Chrome browser session from the command line, you can use a CLI tool that maintains a persistent CDP connection. This allows you to execute commands like open, click, fill, and eval directly against a real browser instance.

Can I control an authenticated Chromium session using an MCP client?

Yes, you can control an authenticated Chromium session using an MCP client. The automation tool maintains a persistent browser session via a daemon, allowing AI models to drive multi-step interactions across complex sites that require login.

What is the best way to interact with elements in a headless browser for SPA rendering?

The best way to interact with elements in a headless browser for SPA rendering is by using snapshot-based interaction. This method obtains element references via the accessibility tree, enabling precise actions like click and type against dynamic page content.

Does Chrome DevTools Protocol automation support cross-domain cookies and browser state?

Yes, Chrome DevTools Protocol automation can support cross-domain cookies and browser state. By using site adapters and extension-backed APIs, the automation tool unlocks browser state for end-to-end workflows across different domains.

How do I handle multi-step interactions on complex sites with a single binary?

You can handle multi-step interactions on complex sites with a single binary by launching a daemon that maintains a persistent CDP connection. This setup provides a unified interface for driving actions and evaluating JavaScript across live sessions.

Can I emulate viewport sizes when automating a real Chromium browser?

Yes, you can emulate viewport sizes when automating a real Chromium browser. The automation system supports viewport emulation and provides snapshot refs to ensure interactions are accurately mapped to the rendered page layout.