Vercel Labs 發布用於 AI 代理的無頭瀏覽器自動化 CLI

Vercel Labs 發布用於 AI 代理的無頭瀏覽器自動化 CLI

Hacker News·

Vercel Labs 推出了名為 'agent-browser' 的新工具,這是一個使用 Rust 開發並支援 Node.js 的命令列介面 (CLI)。該工具專為無頭瀏覽器自動化而設計,旨在賦予 AI 代理更強大的能力。

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

To see all available qualifiers, see our documentation.

Browser automation CLI for AI agents

License

Uh oh!

There was an error while loading. Please reload this page.

vercel-labs/agent-browser

Folders and files

Latest commit

History

Repository files navigation

agent-browser

Headless browser automation CLI for AI agents. Fast Rust CLI with Node.js fallback.

Installation

npm (recommended)

From Source

Linux Dependencies

On Linux, install system dependencies:

Quick Start

Traditional Selectors (also supported)

Commands

Core Commands

Get Info

Check State

Find Elements (Semantic Locators)

Actions: click, fill, check, hover, text

Examples:

Wait

Load states: load, domcontentloaded, networkidle

Mouse Control

Browser Settings

Cookies & Storage

Network

Tabs & Windows

Frames

Dialogs

Debug

Navigation

Setup

Sessions

Run multiple isolated browser instances:

Each session has its own:

Snapshot Options

The snapshot command supports filtering to reduce output size:

Options

Selectors

Refs (Recommended for AI)

Refs provide deterministic element selection from snapshots:

Why use refs?

CSS Selectors

Text & XPath

Semantic Locators

Agent Mode

Use --json for machine-readable output:

Optimal AI Workflow

Headed Mode

Show the browser window for debugging:

This opens a visible browser window instead of running headless.

Architecture

agent-browser uses a client-daemon architecture:

The daemon starts automatically on first command and persists between commands for fast subsequent operations.

Browser Engine: Uses Chromium by default. The daemon also supports Firefox and WebKit via the Playwright protocol.

Platforms

Usage with AI Agents

Just ask the agent

The simplest approach - just tell your agent to use it:

The --help output is comprehensive and most agents can figure it out from there.

AGENTS.md / CLAUDE.md

For more consistent results, add to your project or global instructions file:

Claude Code Skill

For Claude Code, a skill provides richer context:

Or download:

License

Apache-2.0

About

Browser automation CLI for AI agents

Resources

License

Uh oh!

There was an error while loading. Please reload this page.

Stars

Watchers

Forks

Releases

Packages

  0

Contributors

  3

Uh oh!

There was an error while loading. Please reload this page.

Languages

Footer

Footer navigation

Hacker News

相關文章

  1. 適用於 AI 代理的無頭瀏覽器自動化 CLI (Rust)

    3 個月前

  2. 代理技能:用於 React 和 Next.js 工作流程的 AI 代理

    3 個月前

  3. Show HN:Webctl – 基於 CLI 的瀏覽器自動化,適用於代理而非 MCP

    3 個月前

  4. Show HN:Smooth CLI – 適用於 AI 代理的 token 高效瀏覽器

    3 個月前

  5. Vercel 推出 Skills — 內建 React 最佳實踐的「AI 代理的 npm」

    3 個月前