Taobao Product Research

淘宝产品调研和数据采集工具。用于采集淘宝商品信息，包括主图、标题、价格、销量、评价数、店铺等，并生成包含图片的Excel报告。使用场景：(1) 市场调研和竞品分析，(2) 产品价格监控，(3) 销量数据统计，(4) 产品信息批量采集。当用户需要采集淘宝产品数据、进行电商调研、获取商品信息时触发此Skill。

Security Scan

The skill is internally consistent with its description: it uses Playwright to browse Taobao, requires manual login, saves session data and images locally, and produces Excel reports—there are no unexplained credential requests or network exfiltration in the shipped code or SKILL.md.

Purpose & Capability

Name/description (Taobao product research) match the implementation: the code performs browser automation, extracts product fields, downloads images, and writes an Excel file. Declared dependencies (playwright, exceljs, axios) are appropriate for these tasks.

Instruction Scope

SKILL.md instructs running the provided Node script, performing a manual Taobao login in a launched browser, and saving browser state. Instructions do not request unrelated files or external endpoints. Note: the tool takes screenshots and writes debug_*.png and browser_data (login/session data) to disk—these can contain sensitive page content and cookies.

Install Mechanism

There is no automated install spec in the registry metadata; the README instructs running npm install in the scripts folder. That will fetch Playwright and other npm packages from the public registry; Playwright's install will download browser binaries. This is expected for a Playwright-based scraper but means network fetch and binary downloads during install.

Credentials

The skill does not request environment variables, secrets, or unrelated credentials. It stores login/session data locally (browser_data) as part of normal browser automation; no evidence that it exfiltrates credentials or asks for unrelated service keys.

Persistence & Privilege

The skill persists browser profile/login data under its output directory (browser_data). always:false (not force-included). Persisting a logged-in session is expected for convenience but increases the sensitivity of files written by the skill; the skill does not modify other skills or system-wide settings.

Assessment

This skill appears to do what it says (automated browsing of Taobao and Excel report generation). Before installing or running it: (1) Review the script if you want assurance—it's shipped in clear text; (2) run it in a sandboxed or disposable environment if you don't want your Taobao login persisted in the skill folder; (3) be aware npm install will download Playwright and browser binaries from the network—only proceed if you trust those packages; (4) the tool saves screenshots, images, and browser_data (cookies/session) locally—treat these files as sensitive; (5) respect Taobao's terms of service and rate limits to avoid account blocks. If you want extra caution, run the script with a throwaway account or inspect/modify the code to change where browser data and screenshots are written.

淘宝产品调研工具

淘宝产品数据采集和调研工具，支持按关键词、价格区间筛选，自动提取产品信息并生成Excel报告。

功能特性

• 🔍 关键词搜索 - 支持任意关键词搜索
• 💰 价格筛选 - 支持价格区间过滤
• 📊 数据提取 - 自动提取标题、价格、销量、评价数、店铺
• 🖼️ 图片采集 - 下载产品主图并嵌入Excel
• 🔗 链接生成 - 生成可点击的商品链接
• 📁 Excel输出 - 生成格式化的Excel报告

使用方法

命令行方式

cd skills/taobao-product-research/scripts
node taobao_research.js <关键词> <价格区间> <数量> [输出目录]

参数说明:

• 关键词 - 搜索关键词，如 "AI机器人"
• 价格区间 - 价格范围，格式 "min-max"，如 "0-169"
• 数量 - 需要采集的产品数量
• 输出目录 - (可选) 输出文件保存路径

示例:

# 基础用法
node taobao_research.js "AI机器人" "0-169" 15

# 指定输出目录
node taobao_research.js "智能手表" "100-500" 20 ./output

# 高价产品调研
node taobao_research.js "无人机" "1000-5000" 10

程序化调用

const { TaobaoResearcher } = require('./taobao_research.js');

const researcher = new TaobaoResearcher({
    keyword: 'AI机器人',
    priceMin: 0,
    priceMax: 169,
    maxItems: 15,
    outputDir: './research_output'
});

researcher.run().then(result => {
    console.log('完成:', result);
});

首次使用

1. 安装依赖

   cd skills/taobao-product-research/scripts
   npm install
   

2. 登录淘宝

   • 首次运行时会自动打开浏览器
   • 手动完成淘宝登录
   • 按回车键继续
   • 登录状态会保存在 browser_data 目录，后续无需重复登录

输出文件

生成的Excel文件包含以下字段:

字段	说明
序号	排名
主图	产品图片（嵌入）
标题	产品名称
价格	实际售价
销量	付款人数
评价数	评价数量
店铺	店铺名称
链接	可点击的商品链接

文件命名格式: taobao_<关键词>_<价格区间>_<时间戳>.xlsx

注意事项

1. 反爬机制 - 淘宝有反爬机制，建议:

   • 控制采集频率（脚本已内置随机延迟）
   • 单次采集数量不宜过多（建议10-20个）
   • 不同价格区间之间间隔10秒以上

2. 登录状态 - 需要保持淘宝登录状态，如遇到登录失效请重新登录

3. 网络环境 - 确保网络稳定，避免因网络问题导致数据提取失败

4. 价格筛选 - 脚本会双重验证价格，确保在指定范围内

故障排查

问题: 无法提取数据

• 检查是否已登录淘宝
• 查看 debug_*.png 截图文件
• 可能是页面结构变化，需要更新选择器

问题: 图片下载失败

• 检查网络连接
• 图片URL可能已过期，不影响其他数据

问题: 评价数获取失败

• 部分商品可能没有评价
• 详情页加载超时，已设置5秒等待时间

依赖项

• playwright - 浏览器自动化
• xlsx - Excel文件处理
• exceljs - Excel图片嵌入

文件结构

skills/taobao-product-research/
├── SKILL.md
└── scripts/
    ├── package.json
    └── taobao_research.js

运行后生成的文件:

taobao_research/
├── browser_data/          # 浏览器数据（登录状态）
├── images/               # 下载的产品图片
│   ├── 0-169/
│   └── 169-349/
└── taobao_*.xlsx        # Excel报告