@SunNeverSetsX: 分享我珍藏的 4 个让 Agent Loop 跑得更快、更好的基础设施,一定要让你的 Agent 用上! 1. http://agent-browser.dev 让 Agent 在真实浏览器里点击、验证 UI 变更,实现真正的自闭环测试 …
摘要
agent-browser is a CLI tool for browser automation designed for AI agents, using compact text output and ref-based element selection to minimize token usage. The post also highlights three other tools—portless, emulate, and ai-cli—for improving agent loop efficiency.
查看缓存全文
缓存时间: 2026/06/23 06:01
分享我珍藏的 4 个让 Agent Loop 跑得更快、更好的基础设施,一定要让你的 Agent 用上!
-
http://agent-browser.dev 让 Agent 在真实浏览器里点击、验证 UI 变更,实现真正的自闭环测试
-
http://portless.sh 解决端口冲突问题,还完美支持 git worktree
-
http://emulate.dev 本地模拟 Stripe、GitHub 等第三方 API,让 Agent Loop 零成本、零依赖、结果完全确定
-
http://ai-cli.dev 通过简单的 CLI 命令直接生成图像和视频
agent-browser | Browser Automation for AI
Source: https://agent-browser.dev/ Browser automation CLI designed for AI agents. Compact text output minimizes context usage. 100% native Rust.
npm install -g agent-browser # all platforms
brew install agent-browser # macOS
agent-browser install # Download Chrome (first time)
# or try without installing
npx agent-browser open example.com
Features#
- Agent-first: Compact text output uses fewer tokens than JSON, designed for AI context efficiency
- Ref-based: Snapshot returns accessibility tree with refs for deterministic element selection
- Complete: 50+ commands for navigation, forms, screenshots, network, storage, files, tabs, frames, and debugging
- Observable:Video recording,streaming,debugging,profiler, anddiffingtools are built in
- Modern apps:Network control,React & Web Vitals,init scripts, andNext.js + Vercelworkflows have first-class docs
- Stateful:Sessions, profiles, auth state, cookies, storage, proxy, and security controls support long-running agents
- Cross-platform: macOS, Linux, Windows with native binaries
Works with#
Claude Code, Cursor, GitHub Copilot, OpenAI Codex, Google Gemini, opencode, and any agent that can run shell commands.
Example#
# Navigate and get snapshot
agent-browser open example.com
agent-browser snapshot -i
# Output:
# - heading "Example Domain" [ref=e1]
# - link "More information..." [ref=e2]
# Interact using refs
agent-browser click @e2
agent-browser screenshot page.png
agent-browser close
Why refs?#
Thesnapshotcommand returns a compact accessibility tree where each element has a unique ref like@e1,@e2. This provides:
- Context-efficient: Text output uses ~200-400 tokens vs ~3000-5000 for full DOM
- Deterministic: Ref points to exact element from snapshot
- Fast: No DOM re-query needed
- AI-friendly: LLMs parse text output naturally
Architecture#
Client-daemon architecture for optimal performance:
- Rust CLI: Parses commands, communicates with daemon
- Native Daemon: Pure Rust daemon using direct CDP, manages Chrome via Chrome DevTools Protocol
Daemon starts automatically and persists between commands.
Platforms#
Native Rust binaries for macOS (ARM64, x64), Linux (ARM64, x64), and Windows (x64).
相似文章
@cxjwin: 发现一个很实用的网站,专门收集各种现成的 Agent Loop 模板。 核心理念就是:别再一直手动 prompt 了,设计好 loop 让 agent 自己跑完任务。 里面有 Test Until Green、Fix CI Until G…
分享了一个收集现成 Agent Loop 模板的实用网站,核心理念是让 AI agent 自动完成任务,包含 Test Until Green 等常用 loop,适合配合 Claude Code 和 Cursor 使用。
@GitHub_Daily: 让 AI Agent 自动化操作浏览器或抓数据,经常被各种反爬机制拦截,遇到验证码、人机验证直接卡死。 最近 BrowserAct 团队开源了一个 Skill,专为 AI Agent 设计的浏览器自动化命令行工具。 提供三层反封锁机制,从…
BrowserAct 团队开源了一个专为 AI Agent 设计的浏览器自动化命令行工具,提供三层反封锁机制(指纹伪装、验证码破解、人类接管),支持多浏览器并行、账户隔离,并优化了输出格式以节省Token。
@LukeberryPi:如何让 AI 代理控制你的浏览器?— Playwright MCP、browser-use、agentbrowser,还有哪些?
一条推文探讨如何让 AI 代理控制浏览器,列举了 Playwright MCP、browser-use、agentbrowser 等工具,并询问还有哪些替代方案。
@omarsar0:这是一款很棒的产品。如果你正在构建个人智能体,并需要可靠、快速的浏览器使用工具,强烈推荐……
Gumloop 推出 Agent Browsers:在云端沙箱中为 AI 智能体提供可复用的浏览器会话,支持 1Password 凭证管理、会话回放和隐蔽代理,并将任务保存为可每周重复执行的配方。
@shao__meng: Chrome DevTools for Agents 1.0 正式发布 https://developer.chrome.com/blog/devtools-for-agents-v1… 它在真实浏览器中观察行为、检查输出,让 Agent…
Chrome DevTools for Agents 1.0 正式发布,为AI编码Agent提供实时浏览器调试能力,支持MCP服务器、CLI和Agent技能三种接入方式,包含Lighthouse审计、模拟、扩展调试等七大能力。