chrome-use

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

chrome-use

chrome-use

Fast browser automation CLI for AI agents. Chrome/Chromium via CDP with accessibility-tree snapshots and compact
@eN
element refs.
Install / self-heal: if the
chrome-use
(or
abs
) command is missing — e.g. a command errors with "command not found" — install it from the GitHub Release (no npm, no token), then retry. Do NOT fall back to other browser tools:
bash
curl -fsSL https://raw.githubusercontent.com/leeguooooo/chrome-use/main/install.sh | sh
面向AI Agent的快速浏览器自动化CLI工具。通过CDP控制Chrome/Chromium,支持无障碍树快照和简洁的
@eN
元素引用。
安装/自修复: 如果
chrome-use
(或
abs
)命令缺失——例如命令报错“command not found”——请从GitHub Release安装(无需npm,无需令牌),然后重试。请勿切换到其他浏览器工具:
bash
curl -fsSL https://raw.githubusercontent.com/leeguooooo/chrome-use/main/install.sh | sh

Start here

入门指南

This file is a discovery stub, not the usage guide. Before running any
chrome-use
command, load the actual workflow content from the CLI:
bash
chrome-use skills get core             # start here — workflows, common patterns, troubleshooting
chrome-use skills get core --full      # include full command reference and templates
The CLI serves skill content that always matches the installed version, so instructions never go stale. The content in this stub cannot change between releases, which is why it just points at
skills get core
.
本文档是一个发现存根,而非使用指南。在运行任何
chrome-use
命令之前,请从CLI加载实际的工作流内容:
bash
chrome-use skills get core             # 从这里开始——工作流、常见模式、故障排查
chrome-use skills get core --full      # 包含完整命令参考和模板
CLI提供的技能内容始终与已安装版本匹配,因此说明不会过时。本存根中的内容在版本间不会更改,这就是为什么它只指向
skills get core
的原因。

Route to a specialized skill by symptom

根据场景选择专用技能

Load a specialized guide when the task falls outside plain browser web pages. Match on the situation you're in, then run the command — the binary serves the full, version-matched content:
What you're hittingRun
An element is clearly on screen but snapshot/find returns no
@ref
(canvas/WebGL/game/map)
chrome-use skills get canvas
Mock an API response, rewrite request headers, block a URL, record HAR
chrome-use skills get network
Debug React renders/state, or measure LCP/CLS/INP
chrome-use skills get react
Drive the user's real, already-logged-in Chrome (reuse the session)
chrome-use skills get real-chrome
Parallel sessions, multiple accounts, recover a stuck tab
chrome-use skills get sessions
Turn manual checks into a re-runnable regression suite
chrome-use skills get test
Electron desktop apps (VS Code, Slack, Discord, Figma, …)
chrome-use skills get electron
Slack workspace automation
chrome-use skills get slack
Exploratory testing / QA / bug hunt
chrome-use skills get dogfood
chrome-use inside Vercel Sandbox microVMs
chrome-use skills get vercel-sandbox
AWS Bedrock AgentCore cloud browsers
chrome-use skills get agentcore
Run
chrome-use skills list
to see everything available on the installed version.
当任务超出普通网页范围时,请加载专用指南。根据你的当前场景匹配,然后运行命令——二进制文件会提供完整的、版本匹配的内容:
遇到的情况运行命令
元素明明在屏幕上,但快照/查找未返回
@ref
(canvas/WebGL/游戏/地图)
chrome-use skills get canvas
模拟API响应、重写请求头、阻止URL、录制HAR
chrome-use skills get network
调试React渲染/状态,或测量LCP/CLS/INP
chrome-use skills get react
控制用户已登录的真实Chrome(复用会话)
chrome-use skills get real-chrome
并行会话、多账户、恢复卡住的标签页
chrome-use skills get sessions
将手动检查转换为可重复运行的回归测试套件
chrome-use skills get test
Electron桌面应用(VS Code、Slack、Discord、Figma等)
chrome-use skills get electron
Slack工作区自动化
chrome-use skills get slack
探索性测试/QA/漏洞排查
chrome-use skills get dogfood
在Vercel Sandbox微虚拟机中使用chrome-use
chrome-use skills get vercel-sandbox
AWS Bedrock AgentCore云浏览器
chrome-use skills get agentcore
运行
chrome-use skills list
查看已安装版本中所有可用的技能。

Why chrome-use

为什么选择chrome-use

  • Fast native Rust CLI, not a Node.js wrapper
  • Works with any AI agent (Cursor, Claude Code, Codex, Continue, Windsurf, etc.)
  • Chrome/Chromium via CDP with no Playwright or Puppeteer dependency
  • Accessibility-tree snapshots with element refs for reliable interaction
  • Sessions, authentication vault, state persistence, video recording
  • Specialized skills for Electron apps, Slack, exploratory testing, cloud providers
  • 基于Rust的原生快速CLI,而非Node.js包装器
  • 兼容任意AI Agent(Cursor、Claude Code、Codex、Continue、Windsurf等)
  • 通过CDP控制Chrome/Chromium,无需依赖Playwright或Puppeteer
  • 支持无障碍树快照和元素引用,实现可靠交互
  • 会话管理、认证存储、状态持久化、视频录制
  • 针对Electron应用、Slack、探索性测试、云服务商的专用技能

Observability Dashboard

可观测性仪表板

The dashboard runs independently of browser sessions on port 4848 and can also be opened through a proxied or forwarded URL such as
https://dashboard.chrome-use.localhost
. Agents should stay on the dashboard origin: session tabs, status, and stream traffic are proxied internally, so session ports do not need to be exposed.
仪表板独立于浏览器会话运行在4848端口,也可通过代理或转发URL(如
https://dashboard.chrome-use.localhost
)打开。Agent应停留在仪表板源:会话标签页、状态和流量都在内部代理,因此无需暴露会话端口。