bb-browser 0.4.4 → 0.5.0

package/README.md

@@ -16,154 +16,168 @@
 
 ---
 
-You're already logged into Gmail, Twitter, Reddit, Xiaohongshu — bb-browser lets AI agents use that. Not by stealing cookies or faking fingerprints. By **being** the browser.
+You're already logged into Twitter, Reddit, YouTube, Zhihu, Bilibili, LinkedIn, GitHub — bb-browser lets AI agents **use that directly**.
 
+```bash
+bb-browser site twitter/search "AI agent"       # search tweets
+bb-browser site zhihu/hot                        # trending on Zhihu
+bb-browser site arxiv/search "transformer"       # search papers
+bb-browser site eastmoney/stock "茅台"            # real-time stock quote
+bb-browser site boss/search "AI engineer"        # search jobs
+bb-browser site wikipedia/summary "Python"       # Wikipedia summary
+bb-browser site youtube/transcript VIDEO_ID      # full transcript
+bb-browser site stackoverflow/search "async"     # search SO questions
 ```
-AI Agent (Claude Code, Codex, etc.)
-       │ CLI commands
-       ▼
-bb-browser CLI ──HTTP──▶ Daemon ──SSE──▶ Chrome Extension
-                                              │
-                                              ▼ chrome.debugger (CDP)
-                                         Your Real Browser
-                                    (logged-in sites, cookies, sessions)
-```
 
-## Why
+**103 commands across 36 platforms.** All using your real browser's login state. [Full list →](https://github.com/epiral/bb-sites)
+
+## The idea
+
+The internet was built for browsers. AI agents have been trying to access it through APIs — but 99% of websites don't offer one.
 
-Every scraping tool tries to **pretend** it's a browser. bb-browser **is** the browser.
+bb-browser flips this: **instead of forcing websites to provide machine interfaces, let machines use the human interface directly.** The adapter runs `eval` inside your browser tab, calls `fetch()` with your cookies, or invokes the page's own webpack modules. The website thinks it's you. Because it **is** you.
 
 | | Playwright / Selenium | Scraping libs | bb-browser |
 |---|---|---|---|
 | Browser | Headless, isolated | No browser | Your real Chrome |
 | Login state | None, must re-login | Cookie extraction | Already there |
-| Anti-bot | Detected easily | Cat-and-mouse game | Invisible — it IS the user |
-| Internal sites | Need VPN/proxy setup | Can't reach | If you can see it, so can the agent |
+| Anti-bot | Detected easily | Cat-and-mouse | Invisible — it IS the user |
+| Complex auth | Can't replicate | Reverse engineer | Page handles it itself |
 
-## What it can do
+## Quick Start
 
-### Browser Automation
+### Install
 
 ```bash
-bb-browser open https://example.com
-bb-browser snapshot -i           # interactive elements only
-bb-browser click @0              # click by ref
-bb-browser fill @2 "hello"       # fill input
-bb-browser press Enter
-bb-browser screenshot
+npm install -g bb-browser
 ```
 
-### Authenticated Fetch
+### Chrome Extension
+
+1. Download from [Releases](https://github.com/epiral/bb-browser/releases/latest)
+2. Unzip → `chrome://extensions/` → Developer Mode → Load unpacked
 
-Like `curl`, but with your browser's login state. No API keys, no tokens.
+### Use
 
 ```bash
-# Reddit — you're logged in, just fetch
-bb-browser fetch https://www.reddit.com/api/me.json
+bb-browser site update        # pull community adapters
+bb-browser site recommend     # see which adapters match your browsing habits
+bb-browser site zhihu/hot     # go
+```
 
-# Any website's internal API — the browser handles auth
-bb-browser fetch https://api.example.com/user/profile --json
+### MCP (Claude Code / Cursor)
+
+```json
+{
+  "mcpServers": {
+    "bb-browser": {
+      "command": "npx",
+      "args": ["-y", "bb-browser", "--mcp"]
+    }
+  }
+}
 ```
 
-### Network Capture
+## 36 platforms, 103 commands
+
+Community-driven via [bb-sites](https://github.com/epiral/bb-sites). One JS file per command.
 
-See what any website sends and receives — request headers, bodies, response data. Like Chrome DevTools Network tab, but from CLI.
+| Category | Platforms | Commands |
+|----------|-----------|----------|
+| **Search** | Google, Baidu, Bing, DuckDuckGo, Sogou WeChat | search |
+| **Social** | Twitter/X, Reddit, Weibo, Xiaohongshu, Jike, LinkedIn, Hupu | search, feed, thread, user, notifications, hot |
+| **News** | BBC, Reuters, 36kr, Toutiao, Eastmoney | headlines, search, newsflash, hot |
+| **Dev** | GitHub, StackOverflow, HackerNews, CSDN, cnblogs, V2EX, Dev.to, npm, PyPI, arXiv | search, issues, repo, top, thread, package |
+| **Video** | YouTube, Bilibili | search, video, transcript, popular, comments, feed |
+| **Entertainment** | Douban, IMDb, Genius, Qidian | movie, search, top250 |
+| **Finance** | Xueqiu, Eastmoney, Yahoo Finance | stock, hot stocks, feed, watchlist, search |
+| **Jobs** | BOSS Zhipin, LinkedIn | search, detail, profile |
+| **Knowledge** | Wikipedia, Zhihu, Open Library | search, summary, hot, question |
+| **Shopping** | SMZDM | search deals |
+| **Tools** | Youdao, GSMArena, Product Hunt, Ctrip | translate, phone specs, trending products |
+
+## 10 minutes to add any website
 
 ```bash
-bb-browser network requests --filter "api.example.com" --with-body --json
-# → Full request headers (including auth/signing), full response body
-# → See exactly how a website's API works — then build an adapter for it
+bb-browser guide    # full tutorial
 ```
 
-### Site Adapters
+Tell your AI agent: *"turn XX website into a CLI"*. It reads the guide, reverse-engineers the API with `network --with-body`, writes the adapter, tests it, and submits a PR. All autonomously.
+
+Three tiers of adapter complexity:
+
+| Tier | Auth method | Example | Time |
+|------|-------------|---------|------|
+| **1** | Cookie (fetch directly) | Reddit, GitHub, V2EX | ~1 min |
+| **2** | Bearer + CSRF token | Twitter, Zhihu | ~3 min |
+| **3** | Webpack injection / Pinia store | Twitter search, Xiaohongshu | ~10 min |
+
+We tested this: **20 AI agents ran in parallel, each independently reverse-engineered a website and produced a working adapter.** The marginal cost of adding a new website to the agent-accessible internet is approaching zero.
 
-Pre-built commands for popular websites. Community-driven via [bb-sites](https://github.com/epiral/bb-sites).
+## What this means for AI agents
+
+Without bb-browser, an AI agent's world is: **files + terminal + a few APIs with keys.**
+
+With bb-browser: **files + terminal + the entire internet.**
+
+An agent can now, in under a minute:
 
 ```bash
-bb-browser site update                                    # install adapters
-bb-browser site reddit/thread https://reddit.com/r/...    # Reddit discussion tree
-bb-browser site twitter/user yan5xu                        # Twitter profile
-bb-browser site xiaohongshu/feed                           # Xiaohongshu feed
-bb-browser site hackernews/top                             # HN front page
+# Cross-platform research on any topic
+bb-browser site arxiv/search "retrieval augmented generation"
+bb-browser site twitter/search "RAG"
+bb-browser site github search rag-framework
+bb-browser site stackoverflow/search "RAG implementation"
+bb-browser site zhihu/search "RAG"
+bb-browser site 36kr/newsflash
 ```
 
-> Xiaohongshu has request signing (X-s headers). Our adapters call the page's own Vue/Pinia store actions — the page signs the requests itself. Zero reverse engineering needed.
+Six platforms, six dimensions, structured JSON. Faster and broader than any human researcher.
 
-## Install
+## Also a full browser automation tool
 
 ```bash
-npm install -g bb-browser
+bb-browser open https://example.com
+bb-browser snapshot -i                # accessibility tree
+bb-browser click @3                   # click element
+bb-browser fill @5 "hello"            # fill input
+bb-browser eval "document.title"      # run JS
+bb-browser fetch URL --json           # authenticated fetch
+bb-browser network requests --with-body --json  # capture traffic
+bb-browser screenshot                 # take screenshot
 ```
 
-Then load the Chrome extension:
-
-1. `chrome://extensions/` → Enable Developer Mode
-2. "Load unpacked" → select `node_modules/bb-browser/extension/`
-3. Done.
+All commands support `--json` output, `--jq <expr>` for inline filtering, and `--tab <id>` for concurrent multi-tab operations.
 
 ```bash
-bb-browser daemon    # start the daemon
-bb-browser status    # verify connection
+bb-browser site xueqiu/hot-stock 5 --jq '.items[] | {name, changePercent}'
+# {"name":"云天化","changePercent":"2.08%"}
+# {"name":"东芯股份","changePercent":"-7.60%"}
+
+bb-browser site info xueqiu/stock   # view adapter args, example, domain
 ```
 
-## Command Reference
-
-| Category | Command | Description |
-|----------|---------|-------------|
-| **Navigate** | `open <url>` | Open URL |
-| | `back` / `forward` / `refresh` | Navigate |
-| | `close` | Close tab |
-| **Snapshot** | `snapshot` | Full DOM tree |
-| | `snapshot -i` | Interactive elements only |
-| **Interact** | `click <ref>` | Click element |
-| | `fill <ref> <text>` | Clear and fill |
-| | `type <ref> <text>` | Append text |
-| | `hover <ref>` | Hover |
-| | `press <key>` | Keyboard (Enter, Tab, Control+a) |
-| | `scroll <dir> [px]` | Scroll |
-| | `check` / `uncheck <ref>` | Checkbox |
-| | `select <ref> <val>` | Dropdown |
-| **Data** | `get text <ref>` | Element text |
-| | `get url` / `get title` | Page info |
-| | `screenshot [path]` | Screenshot |
-| | `eval "<js>"` | Run JavaScript |
-| **Fetch** | `fetch <url> [--json]` | Authenticated HTTP fetch |
-| **Site** | `site list` | List site adapters |
-| | `site <name> [args]` | Run adapter |
-| | `site update` | Update community adapters |
-| **Network** | `network requests [filter]` | View requests |
-| | `network requests --with-body` | Include headers & body |
-| | `network route "<url>" --abort` | Block requests |
-| | `network clear` | Clear records |
-| **Tab** | `tab` | List tabs |
-| | `tab new [url]` | New tab |
-| | `tab <n>` | Switch tab |
-| **Debug** | `console` / `errors` | Console & JS errors |
-| | `trace start` / `trace stop` | Record user actions |
-| **Daemon** | `daemon` / `stop` / `status` | Manage daemon |
-
-All commands support `--json` for structured output and `--tab <id>` for multi-tab operations.
+## Daemon configuration
+
+The daemon binds to `localhost:19824` by default. You can customize the host with `--host`:
+
+```bash
+bb-browser daemon --host 127.0.0.1    # IPv4 only (fix macOS IPv6 issues)
+bb-browser daemon --host 0.0.0.0      # listen on all interfaces (for Tailscale / ZeroTier remote access)
+```
 
 ## Architecture
 
 ```
-bb-browser/
-├── packages/
-│   ├── cli/          # CLI (TypeScript, argument parsing, HTTP client)
-│   ├── daemon/       # HTTP daemon (SSE bridge, request-response matching)
-│   ├── extension/    # Chrome extension (Manifest V3, chrome.debugger CDP)
-│   └── shared/       # Shared types and protocol definitions
-├── dist/             # Build output (npm publish)
-└── extension/        # Built extension (npm publish)
+AI Agent (Claude Code, Codex, Cursor, etc.)
+       │ CLI or MCP (stdio)
+       ▼
+bb-browser CLI ──HTTP──▶ Daemon ──SSE──▶ Chrome Extension
+                                              │
+                                              ▼ chrome.debugger (CDP)
+                                         Your Real Browser
 ```
 
-| Layer | Tech |
-|-------|------|
-| CLI | TypeScript, zero dependencies |
-| Daemon | Node.js HTTP + SSE |
-| Extension | Chrome MV3 + `chrome.debugger` API |
-| Build | pnpm monorepo + Turborepo + tsup + Vite |
-
 ## License
 
 [MIT](LICENSE)

package/README.zh-CN.md

@@ -16,154 +16,168 @@
 
 ---
 
-你已经登录了 Gmail、Twitter、Reddit、小红书 — bb-browser 让 AI Agent 直接用你的浏览器。不是偷 Cookie，不是伪造指纹。是**成为**浏览器本身。
+你已经登录了微博、知乎、B站、小红书、Twitter、GitHub、LinkedIn — bb-browser 让 AI Agent **直接用你的登录态**。
 
+```bash
+bb-browser site twitter/search "AI agent"       # 搜索推文
+bb-browser site zhihu/hot                        # 知乎热榜
+bb-browser site arxiv/search "transformer"       # 搜论文
+bb-browser site eastmoney/stock "茅台"            # 实时股票行情
+bb-browser site boss/search "AI 工程师"           # 搜职位
+bb-browser site wikipedia/summary "Python"       # 维基百科摘要
+bb-browser site youtube/transcript VIDEO_ID      # YouTube 字幕全文
+bb-browser site stackoverflow/search "async"     # 搜 StackOverflow
 ```
-AI Agent (Claude Code, Codex 等)
-       │ CLI 命令
-       ▼
-bb-browser CLI ──HTTP──▶ Daemon ──SSE──▶ Chrome 扩展
-                                              │
-                                              ▼ chrome.debugger (CDP)
-                                         你的真实浏览器
-                                    (已登录的网站、Cookie、会话)
-```
 
-## 为什么
+**36 个平台，103 个命令，全部用你真实浏览器的登录态。** [完整列表 →](https://github.com/epiral/bb-sites)
+
+## 核心理念
+
+互联网是为浏览器构建的。AI Agent 一直试图通过 API 访问它 — 但 99% 的网站不提供 API。
 
-所有爬虫工具都在**假装**是浏览器。bb-browser **就是**浏览器。
+bb-browser 翻转了这个逻辑：**不是让网站适配机器，而是让机器使用人的界面。** adapter 在你的浏览器 tab 里跑 `eval`，用你的 Cookie 调 `fetch()`，或者直接调用页面的 webpack 模块。网站以为是你在操作。因为**就是你**。
 
 | | Playwright / Selenium | 爬虫库 | bb-browser |
 |---|---|---|---|
 | 浏览器 | 无头、隔离环境 | 没有浏览器 | 你的真实 Chrome |
 | 登录态 | 没有，要重新登录 | 偷 Cookie | 已经在了 |
 | 反爬检测 | 容易被识别 | 猫鼠游戏 | 无法检测 — 它就是用户 |
-| 内部系统 | 需要配 VPN/代理 | 访问不了 | 你能看到的，Agent 都能用 |
+| 复杂鉴权 | 无法复制 | 需要逆向 | 页面自己处理 |
 
-## 能做什么
+## 快速开始
 
-### 浏览器自动化
+### 安装
 
 ```bash
-bb-browser open https://example.com
-bb-browser snapshot -i           # 只看可交互元素
-bb-browser click @0              # 按 ref 点击
-bb-browser fill @2 "hello"       # 填写输入框
-bb-browser press Enter
-bb-browser screenshot
+npm install -g bb-browser
 ```
 
-### 带登录态的 Fetch
+### Chrome 扩展
+
+1. 从 [Releases](https://github.com/epiral/bb-browser/releases/latest) 下载 zip
+2. 解压 → `chrome://extensions/` → 开发者模式 → 加载已解压的扩展程序
 
-像 `curl`，但自动带上浏览器里的 Cookie 和会话。不需要 API key，不需要 token。
+### 使用
 
 ```bash
-# Reddit — 你已经登录了，直接 fetch
-bb-browser fetch https://www.reddit.com/api/me.json
+bb-browser site update        # 拉取社区适配器
+bb-browser site recommend     # 看看哪些和你的浏览习惯匹配
+bb-browser site zhihu/hot     # 开搞
+```
 
-# 任何网站的内部 API — 浏览器自动处理认证
-bb-browser fetch https://api.example.com/user/profile --json
+### MCP 接入（Claude Code / Cursor）
+
+```json
+{
+  "mcpServers": {
+    "bb-browser": {
+      "command": "npx",
+      "args": ["-y", "bb-browser", "--mcp"]
+    }
+  }
+}
 ```
 
-### 网络抓包
+## 36 个平台，103 个命令
 
-查看任何网站发送和接收的完整数据 — 请求头、请求体、响应体。相当于 Chrome DevTools 的 Network 面板，但在命令行里。
+社区驱动，通过 [bb-sites](https://github.com/epiral/bb-sites) 维护。每个命令一个 JS 文件。
+
+| 类别 | 平台 | 命令 |
+|------|------|------|
+| **搜索引擎** | Google、百度、Bing、DuckDuckGo、搜狗微信 | search |
+| **社交媒体** | Twitter/X、Reddit、微博、小红书、即刻、LinkedIn、虎扑 | search、feed、thread、user、notifications、hot |
+| **新闻资讯** | BBC、Reuters、36氪、今日头条、东方财富 | headlines、search、newsflash、hot |
+| **技术开发** | GitHub、StackOverflow、HackerNews、CSDN、博客园、V2EX、Dev.to、npm、PyPI、arXiv | search、issues、repo、top、thread、package |
+| **视频平台** | YouTube、B站 | search、video、transcript、popular、comments、feed |
+| **影音娱乐** | 豆瓣、IMDb、Genius、起点中文网 | movie、search、top250 |
+| **财经股票** | 雪球、东方财富、Yahoo Finance | stock、hot-stock、feed、watchlist、search |
+| **求职招聘** | BOSS直聘、LinkedIn | search、detail、profile |
+| **知识百科** | Wikipedia、知乎、Open Library | search、summary、hot、question |
+| **消费购物** | 什么值得买 | search |
+| **实用工具** | 有道翻译、GSMArena、Product Hunt、携程 | translate、手机参数、热门产品 |
+
+## 10 分钟，CLI 化任何网站
 
 ```bash
-bb-browser network requests --filter "api.example.com" --with-body --json
-# → 完整的请求头（包括签名 header）、完整的响应体
-# → 看清一个网站的 API 是怎么工作的 — 然后为它写 adapter
+bb-browser guide    # 完整教程
 ```
 
-### Site Adapters — 网站 CLI 化
+跟你的 AI Agent 说：*「帮我把 XX 网站 CLI 化」*。它会读 guide，用 `network --with-body` 抓包逆向，写 adapter，测试，然后提 PR 到社区仓库。全程自动。
+
+三种 adapter 复杂度：
+
+| 层级 | 认证方式 | 代表 | 耗时 |
+|------|----------|------|------|
+| **Tier 1** | Cookie（直接 fetch） | Reddit、GitHub、V2EX | ~1 分钟 |
+| **Tier 2** | Bearer + CSRF token | Twitter、知乎 | ~3 分钟 |
+| **Tier 3** | Webpack 注入 / Pinia store | Twitter 搜索、小红书 | ~10 分钟 |
+
+实测：**20 个 AI Agent 并发运行，每个独立逆向一个网站并产出可用的 adapter。** 将一个新网站纳入 Agent 可访问范围的边际成本趋近于零。
 
-为热门网站预置的命令，社区驱动，通过 [bb-sites](https://github.com/epiral/bb-sites) 维护。
+## 对 AI Agent 意味着什么
+
+没有 bb-browser，AI Agent 的世界是：**文件系统 + 终端 + 少数有 API key 的服务。**
+
+有了 bb-browser：**文件系统 + 终端 + 整个互联网。**
+
+一个 Agent 现在可以在一分钟内：
 
 ```bash
-bb-browser site update                                    # 安装/更新 adapter
-bb-browser site reddit/thread https://reddit.com/r/...    # Reddit 讨论树
-bb-browser site twitter/user yan5xu                        # Twitter 用户资料
-bb-browser site xiaohongshu/feed                           # 小红书推荐 Feed
-bb-browser site hackernews/top                             # HN 首页
+# 跨平台调研任何话题
+bb-browser site arxiv/search "retrieval augmented generation"
+bb-browser site twitter/search "RAG"
+bb-browser site github search rag-framework
+bb-browser site stackoverflow/search "RAG implementation"
+bb-browser site zhihu/search "RAG"
+bb-browser site 36kr/newsflash
 ```
 
-> 小红书有请求签名（X-s header）。我们的 adapter 通过调用页面自己的 Vue/Pinia store action 发请求 — 页面自己签名，零逆向。
+六个平台，六个维度，结构化 JSON。比任何人类研究员都快、都广。
 
-## 安装
+## 同时也是完整的浏览器自动化工具
 
 ```bash
-npm install -g bb-browser
+bb-browser open https://example.com
+bb-browser snapshot -i                # 可访问性树
+bb-browser click @3                   # 点击元素
+bb-browser fill @5 "hello"            # 填写输入框
+bb-browser eval "document.title"      # 执行 JS
+bb-browser fetch URL --json           # 带登录态的 fetch
+bb-browser network requests --with-body --json  # 抓包
+bb-browser screenshot                 # 截图
 ```
 
-加载 Chrome 扩展：
-
-1. 打开 `chrome://extensions/` → 开启开发者模式
-2. 点击「加载已解压的扩展程序」→ 选 `node_modules/bb-browser/extension/`
-3. 完成。
+所有命令支持 `--json` 输出、`--jq <expr>` 内联过滤、和 `--tab <id>` 多标签页并发操作。
 
 ```bash
-bb-browser daemon    # 启动 daemon
-bb-browser status    # 确认连接
+bb-browser site xueqiu/hot-stock 5 --jq '.items[] | {name, changePercent}'
+# {"name":"云天化","changePercent":"2.08%"}
+# {"name":"东芯股份","changePercent":"-7.60%"}
+
+bb-browser site info xueqiu/stock   # 查看 adapter 参数、示例、域名
 ```
 
-## 命令速查
+## Daemon 配置
 
-| 分类 | 命令 | 说明 |
-|------|------|------|
-| **导航** | `open <url>` | 打开 URL |
-| | `back` / `forward` / `refresh` | 导航 |
-| | `close` | 关闭标签页 |
-| **快照** | `snapshot` | 完整 DOM 树 |
-| | `snapshot -i` | 只看可交互元素 |
-| **交互** | `click <ref>` | 点击 |
-| | `fill <ref> <text>` | 清空后填入 |
-| | `type <ref> <text>` | 追加输入 |
-| | `hover <ref>` | 悬停 |
-| | `press <key>` | 按键（Enter, Tab, Control+a）|
-| | `scroll <dir> [px]` | 滚动 |
-| | `check` / `uncheck <ref>` | 复选框 |
-| | `select <ref> <val>` | 下拉框 |
-| **数据** | `get text <ref>` | 元素文本 |
-| | `get url` / `get title` | 页面信息 |
-| | `screenshot [path]` | 截图 |
-| | `eval "<js>"` | 执行 JavaScript |
-| **Fetch** | `fetch <url> [--json]` | 带登录态的 HTTP 请求 |
-| **Site** | `site list` | 列出 adapter |
-| | `site <name> [args]` | 运行 adapter |
-| | `site update` | 更新社区 adapter |
-| **网络** | `network requests [filter]` | 查看请求 |
-| | `network requests --with-body` | 包含请求头和响应体 |
-| | `network route "<url>" --abort` | 拦截请求 |
-| | `network clear` | 清空记录 |
-| **Tab** | `tab` | 列出标签页 |
-| | `tab new [url]` | 新标签页 |
-| | `tab <n>` | 切换标签页 |
-| **调试** | `console` / `errors` | 控制台/JS 错误 |
-| | `trace start` / `trace stop` | 录制用户操作 |
-| **Daemon** | `daemon` / `stop` / `status` | 管理 daemon |
-
-所有命令支持 `--json` 结构化输出和 `--tab <id>` 多标签页操作。
+Daemon 默认绑定 `localhost:19824`，可通过 `--host` 自定义监听地址：
+
+```bash
+bb-browser daemon --host 127.0.0.1    # 仅 IPv4（解决 macOS IPv6 问题）
+bb-browser daemon --host 0.0.0.0      # 监听所有网卡（用于 Tailscale / ZeroTier 跨机器访问）
+```
 
 ## 架构
 
 ```
-bb-browser/
-├── packages/
-│   ├── cli/          # CLI（TypeScript，参数解析，HTTP 客户端）
-│   ├── daemon/       # HTTP Daemon（SSE 桥接，请求响应匹配）
-│   ├── extension/    # Chrome 扩展（Manifest V3，chrome.debugger CDP）
-│   └── shared/       # 共享类型和协议定义
-├── dist/             # 构建产物（npm 发布）
-└── extension/        # 构建好的扩展（npm 发布）
+AI Agent (Claude Code, Codex, Cursor 等)
+       │ CLI 或 MCP (stdio)
+       ▼
+bb-browser CLI ──HTTP──▶ Daemon ──SSE──▶ Chrome 扩展
+                                              │
+                                              ▼ chrome.debugger (CDP)
+                                         你的真实浏览器
 ```
 
-| 层 | 技术栈 |
-|----|--------|
-| CLI | TypeScript，零依赖 |
-| Daemon | Node.js HTTP + SSE |
-| Extension | Chrome MV3 + `chrome.debugger` API |
-| 构建 | pnpm monorepo + Turborepo + tsup + Vite |
-
 ## 许可证
 
 [MIT](LICENSE)

package/dist/chunk-AHGAQEFO.js

@@ -0,0 +1,110 @@
+#!/usr/bin/env node
+
+// packages/cli/src/jq.ts
+function splitTopLevel(input, separator) {
+  const parts = [];
+  let current = "";
+  let depth = 0;
+  let inString = false;
+  for (let i = 0; i < input.length; i++) {
+    const char = input[i];
+    const prev = input[i - 1];
+    if (char === '"' && prev !== "\\") inString = !inString;
+    if (!inString) {
+      if (char === "{" || char === "(" || char === "[") depth++;
+      if (char === "}" || char === ")" || char === "]") depth--;
+      if (depth === 0 && input.slice(i, i + separator.length) === separator) {
+        parts.push(current.trim());
+        current = "";
+        i += separator.length - 1;
+        continue;
+      }
+    }
+    current += char;
+  }
+  if (current.trim()) parts.push(current.trim());
+  return parts;
+}
+function parseLiteral(value) {
+  const trimmed = value.trim();
+  if (trimmed.startsWith('"') && trimmed.endsWith('"')) return JSON.parse(trimmed);
+  if (trimmed === "true") return true;
+  if (trimmed === "false") return false;
+  if (trimmed === "null") return null;
+  return Number(trimmed);
+}
+function getField(value, field) {
+  return value !== null && typeof value === "object" ? value[field] : void 0;
+}
+function applySegment(inputs, expr) {
+  if (expr === ".") return inputs;
+  if (expr.startsWith("select(")) {
+    const match = expr.match(/^select\((.+?)\s*(==|>)\s*(.+)\)$/);
+    if (!match) throw new Error(`\u4E0D\u652F\u6301\u7684 jq \u8868\u8FBE\u5F0F: ${expr}`);
+    const [, leftExpr, op, rightExpr] = match;
+    const expected = parseLiteral(rightExpr);
+    return inputs.filter((item) => {
+      const left = applyExpression([item], leftExpr)[0];
+      return op === "==" ? left === expected : Number(left) > Number(expected);
+    });
+  }
+  if (expr.startsWith("{") && expr.endsWith("}")) {
+    const body = expr.slice(1, -1).trim();
+    if (!body) return inputs.map(() => ({}));
+    const entries = splitTopLevel(body, ",");
+    return inputs.map((item) => {
+      const obj = {};
+      for (const entry of entries) {
+        const colon = entry.indexOf(":");
+        if (colon === -1) {
+          const key = entry.trim().replace(/^\./, "");
+          obj[key] = applyExpression([item], `.${key}`)[0];
+        } else {
+          const key = entry.slice(0, colon).trim();
+          const valueExpr = entry.slice(colon + 1).trim();
+          obj[key] = applyExpression([item], valueExpr)[0];
+        }
+      }
+      return obj;
+    });
+  }
+  if (!expr.startsWith(".")) throw new Error(`\u4E0D\u652F\u6301\u7684 jq \u8868\u8FBE\u5F0F: ${expr}`);
+  let current = inputs;
+  let remaining = expr.slice(1);
+  while (remaining.length > 0) {
+    if (remaining.startsWith("[]")) {
+      current = current.flatMap((item) => Array.isArray(item) ? item : []);
+      remaining = remaining.slice(2);
+    } else if (remaining.startsWith("[")) {
+      const match = remaining.match(/^\[(-?\d+)\]/);
+      if (!match) throw new Error(`\u4E0D\u652F\u6301\u7684 jq \u8868\u8FBE\u5F0F: .${remaining}`);
+      const index = Number(match[1]);
+      current = current.map((item) => {
+        if (!Array.isArray(item)) return void 0;
+        return item[index >= 0 ? index : item.length + index];
+      });
+      remaining = remaining.slice(match[0].length);
+    } else if (remaining.startsWith(".")) {
+      remaining = remaining.slice(1);
+    } else {
+      const match = remaining.match(/^([A-Za-z_][A-Za-z0-9_]*)/);
+      if (!match) throw new Error(`\u4E0D\u652F\u6301\u7684 jq \u8868\u8FBE\u5F0F: .${remaining}`);
+      const field = match[1];
+      current = current.map((item) => getField(item, field));
+      remaining = remaining.slice(field.length);
+    }
+  }
+  return current;
+}
+function applyExpression(inputs, expression) {
+  const segments = splitTopLevel(expression.trim(), "|");
+  return segments.reduce((current, segment) => applySegment(current, segment.trim()), inputs);
+}
+function applyJq(data, expression) {
+  return applyExpression([data], expression).filter((item) => item !== void 0);
+}
+
+export {
+  applyJq
+};
+//# sourceMappingURL=chunk-AHGAQEFO.js.map