AI Agent 原本的联网能力(WebSearch、WebFetch)缺少调度策略和浏览器自动化能力。这个 Agent Skill 补上的是:联网策略 + CDP 浏览器操作 + 站点经验积累。兼容所有支持 SKILL.md 的 Agent(Claude Code、Cursor、Gemini CLI、Codex CLI 等)。 帮我安装这个 skill:https://github.com/eze-is/web-access claude plugin marketplace add https://github.com/eze-is/web-access claude plugin install web-access@web-access --scope user
给 AI Agent 装上完整联网能力的 Skill。
🌐 官网 · 📖 设计详解 · ⚡ 快速安装
AI Agent 原本的联网能力(WebSearch、WebFetch)缺少调度策略和浏览器自动化能力。这个 Agent Skill 补上的是:联网策略 + CDP 浏览器操作 + 站点经验积累。兼容所有支持 SKILL.md 的 Agent(Claude Code、Cursor、Gemini CLI、Codex CLI 等)。
推荐必读:Web Access:一个 Skill,拉满 Agent 联网和浏览器能力 ,完整介绍了 Web-Access Skill 的开发细节与 Agent Skill 设计哲学,帮助你也能写出类似通用、高上限的 Skill
| 能力 | 说明 |
|---|---|
| 联网工具自动选择 | WebSearch / WebFetch / curl / Jina / CDP,按场景自主判断,可任意组合 |
| CDP Proxy 浏览器操作 | 直连用户日常浏览器(Chrome / Edge / Chromium 系),天然携带登录态,支持动态页面、交互操作、视频截帧 |
| 三种点击方式 | /click(JS click)、/clickAt(CDP 真实鼠标事件)、/setFiles(文件上传) |
| 本地浏览器书签/历史检索 | find-url.mjs 跨 Chrome / Edge 查询公网搜不到的目标(内部系统)或用户访问过的页面,支持关键词/时间窗/访问频度排序 |
| 并行分治 | 多目标时分发子 Agent 并行执行,共享一个 Proxy,tab 级隔离 |
| 站点经验积累 | 按域名存储操作经验(URL 模式、平台特征、已知陷阱),跨 session 复用 |
| 媒体提取 | 从 DOM 直取图片/视频 URL,或对视频任意时间点截帧分析 |
v2.5.2 更新:
edge://inspect/#remote-debugging 勾选 "Allow remote debugging for this browser instance" 即可config.env(gitignored,首次运行从模板创建),通过 WEB_ACCESS_BROWSER 固定默认浏览器;多浏览器同时开启 toggle 时 Agent 会询问偏好。也支持单次覆盖 --browser <chrome|edge>--browser <chrome|edge> 限定单一浏览器scripts/find-url.mjs,从本地 Chrome 书签/历史按关键词/时间窗/访问频度定位 URL。典型场景:用户提到组织内部系统("我们的 XX 平台"等公网搜不到的目标)、回查之前访问过但不记得地址的页面、查看最近高频访问网站等(场景感谢 @MVPGFC 在 #60 提出)${CLAUDE_SKILL_DIR} 字符串替换语法,修复 Windows Git Bash 路径转换错误和变量未设置问题(#47 #46)! 内联注入方式一:npx skills 一键安装(推荐)
npx skills add eze-is/web-access
skills CLI 是开源的 Agent Skill 包管理器,自动检测你的 Agent 环境并安装到正确位置。
方式二:让 Agent 自动安装
帮我安装这个 skill:https://github.com/eze-is/web-access
方式三:Plugin 安装(Claude Code)
claude plugin marketplace add https://github.com/eze-is/web-access
claude plugin install web-access@web-access --scope user
方式四:手动
git clone https://github.com/eze-is/web-access ~/.claude/skills/web-access
CDP 模式需要 Node.js 22+ 和浏览器(Chrome / Edge)开启远程调试:
chrome://inspect/#remote-debuggingedge://inspect/#remote-debuggingskill 长期偏好保存在 ${CLAUDE_SKILL_DIR}/config.env(首次运行自动从 config.env.template 创建,gitignored):
# 留空 = 每次启动都询问偏好;设值 = 固定使用该浏览器
WEB_ACCESS_BROWSER=edge
合法值:chrome / edge
临时用别的浏览器(不修改 config.env):
node "${CLAUDE_SKILL_DIR}/scripts/check-deps.mjs" --browser chrome
切换浏览器(proxy 已连接旧的):
pkill -f cdp-proxy.mjs && node "${CLAUDE_SKILL_DIR}/scripts/check-deps.mjs"
环境检查(Agent 运行时会自动完成前置检查,无需手动执行):
node "${CLAUDE_SKILL_DIR}/scripts/check-deps.mjs"
# $CLAUDE_SKILL_DIR 是 skill 加载时自动设置的环境变量
# 手动运行请替换为实际路径,如 ~/.claude/skills/web-access
Proxy 通过 WebSocket 直连浏览器(兼容 chrome://inspect / edge://inspect 方式,无需命令行参数启动),提供 HTTP API:
# 启动(Agent 会自动管理 Proxy 生命周期,无需手动启动)
node "${CLAUDE_SKILL_DIR}/scripts/cdp-proxy.mjs" &
# 页面操作
curl -s -X POST --data-raw 'https://example.com' http://localhost:3456/new # 新建 tab(v2.5.3 起 URL 走 POST body)
curl -s -X POST "http://localhost:3456/eval?target=ID" -d 'document.title' # 执行 JS
curl -s -X POST "http://localhost:3456/click?target=ID" -d 'button.submit' # JS 点击
curl -s -X POST "http://localhost:3456/clickAt?target=ID" -d '.upload-btn' # 真实鼠标点击
curl -s -X POST "http://localhost:3456/setFiles?target=ID" \
-d '{"selector":"input[type=file]","files":["/path/to/file.png"]}' # 文件上传
curl -s "http://localhost:3456/screenshot?target=ID&file=/tmp/shot.png" # 截图
curl -s "http://localhost:3456/scroll?target=ID&direction=bottom" # 滚动
curl -s "http://localhost:3456/close?target=ID" # 关闭 tab
curl -s "http://localhost:3456/health" # 查看状态(含 managedTabs 数量)
Proxy 会自动追踪通过 /new 创建的 tab,闲置 15 分钟后自动关闭,防止 Agent 异常退出时留下孤儿 tab。可通过环境变量 CDP_TAB_IDLE_TIMEOUT(单位毫秒)调整超时时间。
通过浏览器自动化操作社交平台(如小红书)存在账号被平台限流或封禁的风险。强烈建议使用小号进行操作。
安装后直接让 Agent 执行联网任务,skill 自动接管:
Skill = 哲学 + 技术事实,不是操作手册。讲清 tradeoff 让 AI 自己选,不替它推理。
详见 SKILL.md 中的浏览哲学部分。
How to use Microsoft Access - Beginner Tutorial
Kevin Stratvert · 4739K views
How To Access the DARK WEB in 2024 (3 Levels)
NetworkChuck · 7549K views
How to Access Dark Web Safely
Dinjik · 1245K views
“China Blocks Web Access to The New York Times After Article”
“Apple Balks at Higher Web-Access Fees After $1B Video Bet”
“LIGHTON: LightOn Enhances Its Paradigm Platform with Real-Time Web Access Through a Strategic Partnership with Linkup - marketscreener.com — marketscreener.com”
“Iran starts restoring web access after monthslong blackout - Semafor — Semafor”
“How To Access One Drive In Outlook Web Access(OWA) In Microsoft Office 365? Pbks Vs Mi (ThdTL7JwvK) - Mshale — Mshale”
AI
Companies use AI to filter candidates. I just gave candidates AI to choose companies. Career-Ops (career-ops.org, also known as careerops) turns any AI coding CLI into a full job search command center. Instead of manually tracking applications in a spreadsheet, you get an AI-powered pipeline that: Career-ops is agentic: Claude Code navigates career pages with Playwright, evaluates fit by reasoning about your CV vs the job description (not keyword matching), and adapts your resume per listing.
AI
CLI-Anything: Bridging the Gap Between AI Agents and the World's Software 🌐 CLI-Hub: pip install cli-anything-hub then cli-hub install — browse, install, and manage all community-built CLIs. Want to add your own? Open a PR — the hub updates instantly. 🎬 See Demos: Watch AI agents use generated CLIs plus preview, live preview, and trajectory loops to produce real artifacts — CAD builds, 3D scenes, diagrams, gameplay, subtitles, and more.
AI
A self-hosted AI workspace -- meant to be the self-hosted version of the UI experience you get from ChatGPT and Claude. But with more jank and fun. Running on your own hardware, with your own data -- local-first, privacy-first, and no trojan. A full, hover-to-play tour lives on the landing page (docs/index.html). Defaults work out of the box: clone, run, then configure models/search/email inside Settings. Only edit .env for deployment-level overrides like APPBIND, APPPORT, AUTHENABLED, DATABASEURL, or a pre-seeded admin password.
AI
Most AI material teaches in scattered pieces. A paper here, a fine-tuning post there, a flashy agent demo somewhere else. The pieces rarely line up. You ship a chatbot but can't explain its loss curve. You hook a function to an agent but can't say what attention does inside the model that's calling it. This curriculum is the spine. 20 phases, 503 lessons, four languages: Python, TypeScript, Rust, Julia. Linear algebra at one end, autonomous swarms at the other. Every algorithm gets built from raw math first. Backprop. Tokenizer. Attention. Agent loop. By the time PyTorch shows up, you already know what it's doing under the hood. Each lesson runs the same loop: read the problem, derive the math, write the code, run the test, keep the artifact. No five-minute videos, no copy-paste deploys,