@sleepinsummer/agent-browser-cli 0.3.1 → 0.3.3

package/CHANGELOG.md

@@ -1,7 +1,26 @@
+## 未发布
+
+## v0.3.3 - 2026-05-18
+
+- `tabtree` 默认改为 compact 输出，截断长 URL 并省略 `session_key` 以降低 token 消耗；新增 `tabtree --full` 输出完整 URL 和 `session_key`。
+- 优化 `tabtree` 实现，daemon 锁内只复制必要会话字段，锁外完成排序和 JSON 组装。
+- Chrome 扩展 popup 增加 label 唯一性提示：推荐用 CLI 设置 label 以校验当前 daemon 内跨 Profile 唯一性。
+- 平台 npm 包打包时默认对复制后的原生二进制执行 `strip`，减小发布体积；本地 `target/release` 二进制保持不变。
+- 新增 `tabtree` 树形查询命令，支持按 `tab_id`、`profile_id/profile_label`、`browser_id` 过滤，并保留 browser → profile → tab 父子节点；原先临时增加的 `profiles` / `browsers` 摘要命令已移除，统一使用 `tabtree` / `lookup`。
+- 新增 `lookup tab|browser|profile` 反查命令，可由 `tab_id` 反查 `browser_id` / `profile_id` / `profile_label`，或由 `browser_id` 反查所属 profile。
+- 移除扩展默认注入的全局 `alert` / `confirm` / `prompt` 重写，改为 CLI 页面执行期间临时抑制弹窗并在命令结束后恢复原生函数。
+- 新增 `profile-label set|clear`，并在 Chrome 扩展 popup 中支持设置 Profile Label；label 冲突时 CLI 按歧义处理，不参与内部路由主键。
+
 # 更新日志
 
 所有重要变更都会记录在这里。日期使用北京时间自然日。
 
+## v0.3.2 - 2026-05-17
+
+### 修复
+
+- 修复长 URL 页面中 `snapshot` 后立即使用 `@e` 可能误报 `ref expired` 的问题。
+
 ## v0.3.1 - 2026-05-17
 
 ### 新增

package/README.md

@@ -10,7 +10,7 @@
   <a href="https://github.com/sleepinginsummer/agent-browser-cli"><img src="https://img.shields.io/badge/CLI-agentbrowsercli-2ea44f" alt="CLI agentbrowsercli"></a>
   <a href="https://github.com/sleepinginsummer/agent-browser-cli/blob/main/LICENSE"><img src="https://img.shields.io/badge/License-MIT-green" alt="License MIT"></a>
   <a href="https://github.com/sleepinginsummer/agent-browser-cli"><img src="https://img.shields.io/badge/sys-win%2Fmac%2Flinux-0078D6?labelColor=0078D6&color=C0C0C0" alt="sys win/mac/linux"></a>
-  <a href="https://github.com/sleepinginsummer/agent-browser-cli/releases"><img src="https://img.shields.io/badge/release-v0.3.1--beta.1-orange" alt="release v0.3.1-beta.1"></a>
+  <a href="https://github.com/sleepinginsummer/agent-browser-cli/releases"><img src="https://img.shields.io/badge/release-v0.3.3-orange" alt="release v0.3.3"></a>
   <a href="https://github.com/sleepinginsummer/agent-browser-cli/pulls"><img src="https://img.shields.io/badge/PRs-welcome-brightgreen" alt="PRs welcome"></a>
 </p>
 
@@ -53,32 +53,36 @@
 - 若干优化，缩短命令执行时间
 - rust实现cli端
 
-## 性能参考
-
-以下为常驻服务已启动、Chrome 扩展已连接时的实测参考，实际耗时会受页面复杂度、网络、Chrome 状态和返回数据量影响。
-
-| 操作 | 参考耗时 |
-| --- | --- |
-| 打开百度标签页 | 约 `0.10s` |
-| 注入 JS 输入关键词并点击搜索 | 约 `0.27s` |
-| 打开百度并搜索“小猫”合计 | 约 `0.37s` |
-| `scan --tab --text-only` 读取页面文本 | 约 `0.04-0.12s` |
-| `exec 'return document.title'` 注入简单 JS | 约 `0.04-0.12s` |
-| `exec 'return document.body.innerText'` 读取正文 | 多数 `0.04-0.05s`，偶发约 `0.30s` |
-| 查询 DOM 链接列表 | 约 `0.27-0.36s` |
-| `exec --monitor` 页面变化摘要 | 约 `0.72-0.88s` |
-
-一般判断：普通读页面和简单 JS 注入是 `50ms` 级；复杂 DOM 查询主要取决于页面结构和返回数据量，常见约 `300ms`；`--monitor` 会额外生成页面变化摘要，通常接近 `0.8s`。
-
-与原 Python 调用链的参考对比：
-
-| 对比项 | Python 版本 | Rust CLI 版本 |
-| --- | --- | --- |
-| 启动方式 | 每次调用更容易触发 Python 进程、模块加载和连接初始化开销 | CLI 命令复用常驻服务，避免重复初始化浏览器连接 |
-| 简单读页面 / JS 注入 | 通常受进程启动和 Python 调用链影响，延迟更不稳定 | 常见 `0.04-0.12s` |
-| 连续多次调用 | 多次短命令开销更明显 | 更适合 Agent 高频调用 |
-
-该对比只用于说明架构差异带来的性能趋势；具体耗时仍取决于页面复杂度、Chrome 状态和返回数据量。
+## 他能做的事情
+
+1. 自动化测试
+   可以复用真实浏览器环境做页面流程验证、表单提交、按钮点击、跳转检查、登录态页面测试。
+2. 前端页面 Debug
+   可以读取 DOM、执行 JS、查看页面状态、截图确认效果，辅助定位前端交互、渲染和数据问题，对接后端接口。
+3. 页面样式调试
+   可以在真实页面里执行 JS 修改 DOM / CSS，临时验证样式、布局和交互效果，但更偏辅助调试，不是完整设计工具。
+4. 网页数据采集
+   可以读取页面内容、表格、列表、Cookie 和接口相关状态，适合处理需要登录态的页面数据提取。
+5. 浏览器操作脚本化
+   可以把打开页面、切换标签页、执行 JS、截图、上传文件等操作串成脚本，做重复性网页任务。
+6. Agent 辅助操作网页后台
+   适合让 AI Agent 操作管理后台、配置页面、低代码平台、表单系统等已有网页工具。
+7. 页面结构分析
+   可以简化 HTML、识别主要内容区和列表结构，帮助 Agent 更快理解复杂页面。
+8. 安全研究和逆向辅助
+   可以在真实浏览器会话里观察页面行为、执行调试脚本、读取前端状态，辅助分析前端逻辑和接口调用。
+
+## 他的能力
+
+1. 扫描当前 Chrome 标签页，获取页面标题、URL 和标签页 ID。
+2. 切换到指定标签页，复用已有页面和登录态。
+3. 打开新标签页，支持直接访问目标 URL。
+4. 在页面中执行 JavaScript，读取 DOM、表单、状态和页面数据。
+5. 读取当前页面 Cookie，方便处理登录态相关任务。
+6. 调用 Chrome CDP 能力，执行更底层的页面控制。
+7. 截取页面截图，用于视觉检查和页面确认。
+8. 上传本地文件到网页文件选择框。
+9. 操作下拉框、按钮、表单等常见页面元素。
 
 ## 目录结构
 
@@ -156,7 +160,25 @@ agent-browser-cli set-extension-port 18766
 
 Chrome 插件 popup 中也可以修改插件端口并立即重连。插件端口必须和 CLI 配置中的 `extension_port` 一致。
 
+### Profile Label
+
+多 Chrome Profile / 多浏览器实例下，`profile_id` 和 `browser_id` 较长。可以给每个 Chrome Profile 设置短 label，之后用 `--profile <label>` 操作。
+
+```bash
+agent-browser-cli lookup tab <tabId>
+agent-browser-cli lookup browser <browser_id>
+agent-browser-cli profile-label set work --profile <profile_id>
+agent-browser-cli tabs --profile work
+```
+
+也可以在对应 Chrome Profile 的扩展 popup 中设置 Profile Label。label 只作为别名，内部路由仍使用 `browser_id:profile_id:tab_id`；如果当前 daemon 内 label 匹配到多个 profile，CLI 会报歧义。推荐用 CLI 设置 label，因为 CLI 会校验当前 daemon 内跨 Profile 唯一性；popup 是本地便捷入口，不保证跨 Profile 唯一。`tabtree` 默认截断 URL 并省略 `session_key` 以减少 token，需完整字段时加 `--full`。
+
+
+
+
+### 弹窗抑制
 
+扩展不再默认重写页面的 `alert` / `confirm` / `prompt`。只有 CLI 执行页面脚本命令期间会临时抑制原生弹窗，命令结束后恢复，避免长期污染业务页面全局函数。
 
 ## 快速自检
 
@@ -185,6 +207,14 @@ README 只保留快速入口；完整命令和浏览器操作 SOP 见 [skills/ag
 
 ```bash
 agent-browser-cli tabs
+agent-browser-cli tabtree
+agent-browser-cli tabtree --full
+agent-browser-cli tabtree --profile <profile_label>
+agent-browser-cli tabtree --tab <tabId>
+agent-browser-cli lookup tab <tabId>
+agent-browser-cli lookup browser <browser_id>
+agent-browser-cli profile-label set work --profile <profile_id>
+agent-browser-cli profile-label clear --profile <profile_id>
 ```
 
 ## 更新
@@ -210,6 +240,7 @@ agent-browser-cli stop
 然后按需清理：
 
 ```bash
+npm uninstall -g @sleepinsummer/agent-browser-cli
 rm -f .agent-browser-cli.log .agent-browser-cli.lock
 rm -rf ~/.agents/skills/agent-browser-cli
 ```

package/README_EN.md

@@ -53,32 +53,36 @@ Please read https://github.com/sleepinginsummer/agent-browser-cli/blob/main/AI_I
 - Includes several optimizations to reduce command execution time.
 - Rust implementation for the CLI side.
 
-## Performance Reference
-
-The following numbers are measured with the long-lived service already running and the Chrome extension already connected. Actual latency depends on page complexity, network conditions, Chrome state, and response size.
-
-| Operation | Reference Latency |
-| --- | --- |
-| Open a Baidu tab | About `0.10s` |
-| Inject JS to enter a keyword and submit search | About `0.27s` |
-| Open Baidu and search “小猫” end-to-end | About `0.37s` |
-| `scan --tab --text-only` to read page text | About `0.04-0.12s` |
-| `exec 'return document.title'` for simple JS | About `0.04-0.12s` |
-| `exec 'return document.body.innerText'` to read body text | Mostly `0.04-0.05s`, occasional `0.30s` |
-| Query DOM link lists | About `0.27-0.36s` |
-| `exec --monitor` page-change summary | About `0.72-0.88s` |
-
-Rule of thumb: normal page reads and simple JS injection are around the `50ms` level; complex DOM queries depend on page structure and returned data size, commonly around `300ms`; `--monitor` adds page-change summary work and is usually close to `0.8s`.
-
-Reference comparison with the original Python call chain:
-
-| Item | Python Version | Rust CLI Version |
-| --- | --- | --- |
-| Startup model | Each call is more likely to pay for Python process startup, module loading, and connection initialization | CLI commands reuse the long-lived service and avoid repeated browser connection initialization |
-| Simple page read / JS injection | Usually more affected by process startup and the Python call chain, so latency is less stable | Commonly `0.04-0.12s` |
-| Repeated calls | Overhead is more visible across many short commands | Better suited for high-frequency Agent calls |
-
-This comparison is intended to describe the performance trend caused by the architecture difference. Actual latency still depends on page complexity, Chrome state, and response size.
+## What It Can Do
+
+1. Automated testing
+   It can reuse a real browser environment for page-flow validation, form submission, button clicks, navigation checks, and login-state page testing.
+2. Frontend page debugging
+   It can read the DOM, execute JS, inspect page state, and capture screenshots to help locate frontend interaction, rendering, and data issues, including backend API integration problems.
+3. Page style debugging
+   It can execute JS on real pages to modify DOM / CSS and temporarily validate styles, layout, and interaction effects. It is more of a debugging assistant than a complete design tool.
+4. Web data extraction
+   It can read page content, tables, lists, cookies, and API-related state, making it suitable for extracting data from pages that require login state.
+5. Browser operation scripting
+   It can chain operations such as opening pages, switching tabs, executing JS, taking screenshots, and uploading files into scripts for repetitive web tasks.
+6. Agent-assisted web admin operation
+   It is suitable for letting AI Agents operate existing web tools such as admin consoles, configuration pages, low-code platforms, and form systems.
+7. Page structure analysis
+   It can simplify HTML and identify main content areas and list structures, helping Agents understand complex pages faster.
+8. Security research and reverse-engineering assistance
+   It can observe page behavior in a real browser session, execute debugging scripts, and read frontend state to assist analysis of frontend logic and API calls.
+
+## Its Capabilities
+
+1. Scan current Chrome tabs and get page titles, URLs, and tab IDs.
+2. Switch to a specified tab and reuse existing pages and login state.
+3. Open new tabs and directly visit target URLs.
+4. Execute JavaScript in pages and read DOM, forms, state, and page data.
+5. Read cookies from the current page for login-state related tasks.
+6. Call Chrome CDP capabilities for lower-level page control.
+7. Capture page screenshots for visual inspection and page confirmation.
+8. Upload local files to web file picker inputs.
+9. Operate common page elements such as dropdowns, buttons, and forms.
 
 ## Layout
 
@@ -156,6 +160,10 @@ After manually editing the config file, run `agent-browser-cli restart` so the d
 
 The Chrome extension popup can also update the extension port and reconnect immediately. The popup port must match the CLI `extension_port` config.
 
+### Dialog Suppression
+
+The extension no longer rewrites page `alert` / `confirm` / `prompt` globally. Native dialogs are suppressed only while a CLI page script command is running, then restored after the command finishes.
+
 ## Quick Check
 
 ```bash
@@ -177,12 +185,33 @@ On success, it returns:
 }
 ```
 
+### Profile Label
+
+When multiple Chrome Profiles or browser instances are connected, `profile_id` and `browser_id` are long. You can set a short label for each Chrome Profile and then use `--profile <label>`.
+
+```bash
+agent-browser-cli lookup tab <tabId>
+agent-browser-cli lookup browser <browser_id>
+agent-browser-cli profile-label set work --profile <profile_id>
+agent-browser-cli tabs --profile work
+```
+
+You can also set Profile Label from the extension popup in the corresponding Chrome Profile. The label is only an alias; internal routing still uses `browser_id:profile_id:tab_id`. If a label matches multiple profiles in the current daemon, the CLI reports ambiguity. Prefer setting labels via CLI because it checks uniqueness across currently connected profiles; the popup is a local convenience entry and cannot guarantee cross-profile uniqueness. `tabtree` uses compact output by default; pass `--full` for full URLs and session keys.
+
 ## Common Commands
 
 The README only keeps the quick entry point. For the full command list and browser operation SOP, see [skills/agent-browser-cli/SKILL.md](./skills/agent-browser-cli/SKILL.md).
 
 ```bash
 agent-browser-cli tabs
+agent-browser-cli tabtree
+agent-browser-cli tabtree --full
+agent-browser-cli tabtree --profile <profile_label>
+agent-browser-cli tabtree --tab <tabId>
+agent-browser-cli lookup tab <tabId>
+agent-browser-cli lookup browser <browser_id>
+agent-browser-cli profile-label set work --profile <profile_id>
+agent-browser-cli profile-label clear --profile <profile_id>
 ```
 
 ## Update

package/assets/tmwd_cdp_bridge/background.js

@@ -42,6 +42,7 @@ function withClientIdentity(payload) {
 async function handleExtMessage(msg, sender) {
   if (msg.cmd === 'status') return handleStatus();
   if (msg.cmd === 'setPort') return await handleSetPort(msg);
+  if (msg.cmd === 'setProfileLabel') return await handleSetProfileLabel(msg);
   lastCommandAt = Date.now();
   if (msg.cmd === 'cookies') return await handleCookies(msg, sender);
   if (msg.cmd === 'cdp') return await handleCDP(msg, sender);
@@ -137,6 +138,29 @@ async function handleSetPort(msg) {
   };
 }
 
+function normalizeProfileLabel(label) {
+  const raw = String(label || '').trim();
+  if (!raw) return null;
+  if (raw.length > 40) throw new Error('Profile Label 长度不能超过 40 个字符');
+  if (!/^[A-Za-z0-9_.-]+$/.test(raw)) throw new Error('Profile Label 只能包含英文、数字、-、_、.');
+  return raw;
+}
+
+async function handleSetProfileLabel(msg) {
+  const label = normalizeProfileLabel(msg.label);
+  profileLabel = label;
+  await chrome.storage.local.set({ profileLabel });
+  await sendTabsUpdate();
+  return {
+    ok: true,
+    data: {
+      browserId,
+      profileId,
+      profileLabel
+    }
+  };
+}
+
 chrome.runtime.onMessage.addListener((msg, sender, sendResponse) => {
   handleExtMessage(msg, sender).then(sendResponse);
   return true;
@@ -618,6 +642,87 @@ async function injectContentScriptsIntoExistingTabs() {
   }
 }
 
+function buildDialogSuppressionScript(enabled) {
+  if (!enabled) {
+    return `(() => {
+      const state = window.__TMWD_DIALOG_SUPPRESSION__;
+      if (!state) return;
+      state.count = Math.max(0, Number(state.count || 1) - 1);
+      if (state.count > 0) return;
+      try {
+        window.alert = state.alert;
+        window.confirm = state.confirm;
+        window.prompt = state.prompt;
+      } finally {
+        delete window.__TMWD_DIALOG_SUPPRESSION__;
+      }
+    })()`;
+  }
+  return `(() => {
+    const existing = window.__TMWD_DIALOG_SUPPRESSION__;
+    if (existing) {
+      existing.count = Number(existing.count || 1) + 1;
+      return;
+    }
+    const state = {
+      count: 1,
+      alert: window.alert,
+      confirm: window.confirm,
+      prompt: window.prompt
+    };
+    window.__TMWD_DIALOG_SUPPRESSION__ = state;
+    const toast = (type, msg) => {
+      try { console.log('[TMWD] ' + type + ' suppressed during CLI command:', msg); } catch (_) {}
+      try {
+        const d = document.createElement('div');
+        d.textContent = '[' + type + '] ' + msg;
+        Object.assign(d.style, {
+          position:'fixed', top:'12px', right:'12px', zIndex:'2147483647',
+          background:'#222', color:'#fff', padding:'10px 18px', borderRadius:'8px',
+          fontSize:'14px', maxWidth:'420px', wordBreak:'break-all',
+          boxShadow:'0 4px 16px rgba(0,0,0,.3)', opacity:'1',
+          transition:'opacity .5s', pointerEvents:'none'
+        });
+        (document.body || document.documentElement).appendChild(d);
+        setTimeout(() => { d.style.opacity = '0'; }, 3000);
+        setTimeout(() => { d.remove(); }, 3600);
+      } catch (_) {}
+    };
+    window.alert = function(msg) { toast('alert', msg); };
+    window.confirm = function(msg) { toast('confirm', msg); return true; };
+    window.prompt = function(msg, def) { toast('prompt', msg); return def || null; };
+  })()`;
+}
+
+async function setDialogSuppressionByScripting(tabId, enabled) {
+  try {
+    await chrome.scripting.executeScript({
+      target: { tabId },
+      world: 'MAIN',
+      func: async (script) => await eval(script),
+      args: [buildDialogSuppressionScript(enabled)]
+    });
+    return true;
+  } catch (e) {
+    console.log('[TMWD-WS] dialog suppression scripting failed:', e.message);
+    return false;
+  }
+}
+
+async function setDialogSuppressionByCdp(tabId, enabled) {
+  try {
+    await chrome.debugger.sendCommand({ tabId }, 'Runtime.evaluate', {
+      expression: buildDialogSuppressionScript(enabled),
+      awaitPromise: true,
+      returnByValue: true
+    });
+    return true;
+  } catch (e) {
+    console.log('[TMWD-WS] dialog suppression CDP failed:', e.message);
+    return false;
+  }
+}
+
 // --- Shared page/CDP script builder core ---
 function buildExecScript(code, errorHandler) {
   return `(async () => {
@@ -764,6 +869,7 @@ async function handleWsExec(data) {
   const newTabIds = new Set();
   const onCreated = (tab) => { newTabIds.add(tab.id); };
   chrome.tabs.onCreated.addListener(onCreated);
+  await setDialogSuppressionByScripting(tabId, true);
   try {
     let res;
     try {
@@ -788,9 +894,11 @@ async function handleWsExec(data) {
       const wrappedCode = buildCdpScript(data.code);
       try {
         await chrome.debugger.attach({ tabId }, '1.3');
+        await setDialogSuppressionByCdp(tabId, true);
         const cdpRes = await chrome.debugger.sendCommand({ tabId }, 'Runtime.evaluate', {
           expression: wrappedCode, awaitPromise: true, returnByValue: true
         });
+        await setDialogSuppressionByCdp(tabId, false);
         await chrome.debugger.detach({ tabId });
         if (cdpRes.exceptionDetails) {
           const desc = cdpRes.exceptionDetails.exception?.description || 'CDP Error';
@@ -799,6 +907,7 @@ async function handleWsExec(data) {
           res = cdpRes.result.value;
         }
       } catch (cdpErr) {
+        try { await setDialogSuppressionByCdp(tabId, false); } catch (_) {}
         try { await chrome.debugger.detach({ tabId }); } catch (_) {}
         res = { ok: false, error: { name: 'Error', message: 'CDP fallback failed: ' + cdpErr.message, stack: '' } };
       }
@@ -812,7 +921,7 @@ async function handleWsExec(data) {
       try { const t = await chrome.tabs.get(id); newTabs.push({id: t.id, url: t.url, title: t.title}); } catch (_) {}
     }
     if (res?.ok) {
-      ws.send(JSON.stringify({ type: 'result', id: data.id, result: res.data, newTabs }));
+      ws.send(JSON.stringify({ type: 'result', id: data.id, result: res.data ?? null, newTabs }));
     } else {
       console.log(res);
       ws.send(JSON.stringify({ type: 'error', id: data.id, error: res?.error || 'Unknown error', newTabs }));
@@ -820,6 +929,7 @@ async function handleWsExec(data) {
   } catch (e) {
     ws.send(JSON.stringify({ type: 'error', id: data.id, error: { name: e.name || 'Error', message: e.message || String(e), stack: e.stack || '' } }));
   } finally {
+    await setDialogSuppressionByScripting(tabId, false);
     chrome.tabs.onCreated.removeListener(onCreated);
   }
 }

package/assets/tmwd_cdp_bridge/disable_dialogs.js

@@ -1,24 +1,3 @@
-// Disable alert/confirm/prompt to prevent page JS from blocking extension
-(function() {
-  const _log = console.log.bind(console);
-  function toast(type, msg) {
-    _log('[TMWD] ' + type + ' suppressed:', msg);
-    try {
-      const d = document.createElement('div');
-      d.textContent = '[' + type + '] ' + msg;
-      Object.assign(d.style, {
-        position:'fixed', top:'12px', right:'12px', zIndex:'2147483647',
-        background:'#222', color:'#fff', padding:'10px 18px', borderRadius:'8px',
-        fontSize:'14px', maxWidth:'420px', wordBreak:'break-all',
-        boxShadow:'0 4px 16px rgba(0,0,0,.3)', opacity:'1',
-        transition:'opacity .5s', pointerEvents:'none'
-      });
-      (document.body || document.documentElement).appendChild(d);
-      setTimeout(() => { d.style.opacity = '0'; }, 3000);
-      setTimeout(() => { d.remove(); }, 3600);
-    } catch(e) {}
-  }
-  window.alert = function(msg) { toast('alert', msg); };
-  window.confirm = function(msg) { toast('confirm', msg); return true; };
-  window.prompt = function(msg, def) { toast('prompt', msg); return def || null; };
-})();
+// Deprecated: dialog suppression is no longer injected globally.
+// background.js now enables alert/confirm/prompt suppression only during CLI page execution
+// and restores the native functions after the command finishes.

package/assets/tmwd_cdp_bridge/manifest.json

@@ -23,17 +23,6 @@
     "service_worker": "background.js"
   },
   "content_scripts": [
-    {
-      "matches": [
-        "<all_urls>"
-      ],
-      "js": [
-        "disable_dialogs.js"
-      ],
-      "run_at": "document_start",
-      "all_frames": true,
-      "world": "MAIN"
-    },
     {
       "matches": [
         "<all_urls>"

package/assets/tmwd_cdp_bridge/popup.html

@@ -9,6 +9,7 @@ button{background:#264f78;color:#fff;border:none;padding:4px 12px;cursor:pointer
 button:hover{background:#37699e}
 input{width:86px;background:#252526;color:#d4d4d4;border:1px solid #3c3c3c;border-radius:3px;padding:4px;font:12px monospace}
 label{display:inline-flex;align-items:center;gap:6px;margin-right:6px}
+#profileLabel{width:150px}
 .panel{margin-bottom:8px;padding:6px;background:#252526;border-radius:3px}
 .row{display:flex;align-items:center;gap:6px;margin-top:6px}
 .status{color:#9cdcfe}
@@ -26,6 +27,17 @@ pre{white-space:pre-wrap;word-break:break-all;margin:0;padding:6px;background:#2
   </div>
   <div id="portMsg"></div>
 </div>
+<h3>👤 Profile</h3>
+<div class="panel">
+  <div id="profileStatus" class="status">读取 Profile...</div>
+  <div class="row">
+    <label>Label <input id="profileLabel" type="text" maxlength="40" placeholder="work"></label>
+    <button id="saveProfileLabel">保存</button>
+    <button id="clearProfileLabel">清空</button>
+  </div>
+  <div id="profileMsg"></div>
+  <div class="status">建议用 CLI 设置 label，可校验跨 Profile 唯一性。</div>
+</div>
 <h3>🍪 Cookies</h3>
 <button id="refresh">刷新</button>
 <pre id="out">点击刷新获取 cookies...</pre>

package/assets/tmwd_cdp_bridge/popup.js

@@ -1,8 +1,12 @@
 document.addEventListener('DOMContentLoaded', () => {
   const btn = document.getElementById('refresh');
   const savePortBtn = document.getElementById('savePort');
+  const saveProfileLabelBtn = document.getElementById('saveProfileLabel');
+  const clearProfileLabelBtn = document.getElementById('clearProfileLabel');
   btn.addEventListener('click', fetchCookies);
   savePortBtn.addEventListener('click', savePort);
+  saveProfileLabelBtn.addEventListener('click', saveProfileLabel);
+  clearProfileLabelBtn.addEventListener('click', clearProfileLabel);
   refreshBridgeStatus();
   fetchCookies();
 });
@@ -16,9 +20,49 @@ async function refreshBridgeStatus() {
     const data = resp.data || {};
     portInput.value = data.wsPort || 18765;
     status.textContent = `状态: ${data.wsConnected ? '已连接' : '未连接'} ${data.wsUrl || ''}`;
+    renderProfileStatus(data);
   } catch (e) {
     status.textContent = '状态读取失败: ' + e.message;
     status.className = 'error';
+    const profileStatus = document.getElementById('profileStatus');
+    profileStatus.textContent = 'Profile 读取失败: ' + e.message;
+    profileStatus.className = 'error';
+  }
+}
+
+function renderProfileStatus(data) {
+  const profileStatus = document.getElementById('profileStatus');
+  const profileLabelInput = document.getElementById('profileLabel');
+  const profileId = data.profileId || '-';
+  const browserId = data.browserId || '-';
+  const label = data.profileLabel || '';
+  profileLabelInput.value = label;
+  profileStatus.textContent = `Profile: ${label || '(未设置)'} / ${profileId} / ${browserId}`;
+  profileStatus.className = 'status';
+}
+
+async function saveProfileLabel() {
+  const input = document.getElementById('profileLabel');
+  await setProfileLabel(input.value);
+}
+
+async function clearProfileLabel() {
+  const input = document.getElementById('profileLabel');
+  input.value = '';
+  await setProfileLabel(null);
+}
+
+async function setProfileLabel(label) {
+  const profileMsg = document.getElementById('profileMsg');
+  try {
+    const resp = await chrome.runtime.sendMessage({ cmd: 'setProfileLabel', label });
+    if (!resp?.ok) throw new Error(resp?.error || 'unknown');
+    profileMsg.textContent = `Success: Profile Label ${resp.data?.profileLabel || '已清空'}`;
+    profileMsg.className = 'status';
+    await refreshBridgeStatus();
+  } catch (e) {
+    profileMsg.textContent = '保存失败: ' + e.message;
+    profileMsg.className = 'error';
   }
 }
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sleepinsummer/agent-browser-cli",
-  "version": "0.3.1",
+  "version": "0.3.3",
   "description": "Agent-oriented browser sensing and control CLI backed by a native Rust daemon.",
   "license": "MIT",
   "bin": {
@@ -24,11 +24,11 @@
     "postinstall": "node npm/postinstall.js"
   },
   "optionalDependencies": {
-    "@sleepinsummer/agent-browser-cli-darwin-arm64": "0.3.1",
-    "@sleepinsummer/agent-browser-cli-darwin-x64": "0.3.1",
-    "@sleepinsummer/agent-browser-cli-linux-x64": "0.3.1",
-    "@sleepinsummer/agent-browser-cli-linux-arm64": "0.3.1",
-    "@sleepinsummer/agent-browser-cli-win32-x64": "0.3.1"
+    "@sleepinsummer/agent-browser-cli-darwin-arm64": "0.3.3",
+    "@sleepinsummer/agent-browser-cli-darwin-x64": "0.3.3",
+    "@sleepinsummer/agent-browser-cli-linux-x64": "0.3.3",
+    "@sleepinsummer/agent-browser-cli-linux-arm64": "0.3.3",
+    "@sleepinsummer/agent-browser-cli-win32-x64": "0.3.3"
   },
   "engines": {
     "node": ">=18"

package/skills/agent-browser-cli/SKILL.md

@@ -43,7 +43,15 @@ agent-browser-cli status
 agent-browser-cli doctor
 agent-browser-cli logs --tail 100
 agent-browser-cli tabs
+agent-browser-cli tabtree
+agent-browser-cli tabtree --full
+agent-browser-cli tabtree --profile work
+agent-browser-cli tabtree --tab <tabId>
+agent-browser-cli lookup tab <tabId>
+agent-browser-cli lookup browser <browser_id>
+agent-browser-cli lookup profile work
 agent-browser-cli tabs --profile work
+agent-browser-cli profile-label set work --profile <profile_id>
 agent-browser-cli scan --tabs-only
 agent-browser-cli scan --profile work --tab <tabId> --text-only
 agent-browser-cli open --profile work https://example.com
@@ -81,7 +89,7 @@ agent-browser-cli send-keys --target '@e2' 'Enter'
 agent-browser-cli mouse-click '@e3'
 ```
 
-所有高层操作都支持 `--tab <tabId>`，多 Chrome Profile / 多浏览器实例时还支持 `--profile <profile_id-or-label>` 和 `--browser <browser_id>`。`tabs` 输出会包含 `browser_id`、`profile_id`、`profile_label`、`tab_id`、`session_key`。只传 `--tab` 且存在歧义时，必须补 `--profile` 或 `--browser`，不要猜。`@e` 只在当前 daemon、当前 session_key、最近一次 `snapshot` 内有效。`@e` 只接受 `@e1` 这种带 `@` 的格式。
+所有高层操作都支持 `--tab <tabId>`，多 Chrome Profile / 多浏览器实例时还支持 `--profile <profile_id-or-label>` 和 `--browser <browser_id>`。`tabs` 输出会包含 `browser_id`、`profile_id`、`profile_label`、`tab_id`、`session_key`。`tabtree` 用树形结构列出 browser → profile → tabs，支持 `--tab <tabId>`、`--profile <profile_id-or-label>`、`--browser <browser_id>` 过滤，过滤结果仍保留父子节点；默认 compact 输出会截断 URL 并省略 `session_key`，需要完整字段时用 `tabtree --full`；`lookup tab <tabId>` 可由 tab 反查 `browser_id` / `profile_id` / `profile_label`，`lookup browser <browser_id>` 可反查所属 profile。`profile-label set <label> --profile <profile_id>` 可设置当前 Chrome Profile 的 label 并校验当前 daemon 内唯一性，`profile-label clear --profile <profile_id>` 可清空；popup 设置是本地便捷入口，不保证跨 Profile 唯一；label 只允许英文、数字、`-`、`_`、`.`，当前 daemon 内匹配到多个 profile 时会报歧义，不能猜。只传 `--tab` 且存在歧义时，必须补 `--profile` 或 `--browser`。`@e` 只在当前 daemon、当前 session_key、最近一次 `snapshot` 内有效。`@e` 只接受 `@e1` 这种带 `@` 的格式。
 
 慢页面要把等待和监控分开：
 
@@ -91,6 +99,8 @@ agent-browser-cli click '@e1' --wait-js 'return document.body.innerText.includes
 
 `--wait-js` 负责等慢加载；`--monitor` 只负责操作前后页面 diff，默认关闭。
 
+弹窗处理：扩展默认不改写业务页面的 `alert` / `confirm` / `prompt`。只有 CLI 页面执行命令期间临时抑制弹窗，结束后恢复；如果怀疑页面原生弹窗行为异常，先让用户重载扩展和页面。
+
 ## 端口和扩展
 
 固定 API 端口：