Shopify 网店分类链接采集器 - 自动从导航提取真实分类层级，处理 Ajax 懒加载下拉菜单，一级分类二级分类分别放在不同单元格，输出 CSV。支持 Shopify 多级别分类导航。

This package appears to be a coherent Shopify navigation scraper, but take these precautions before installing or running it: - Review and run the code in an isolated environment (VM/container) first. The scripts will launch a browser and visit many pages on the target site and write files locally. - Install dependencies explicitly: run 'npm install' in the package directory, then 'npx playwright install chromium' (Playwright requires browser binaries). The package-lock references a mirror (npmmirror); you may prefer to use the official registry (registry.npmjs.org) when installing. - Edit the default output path: the scripts default to C:\workspace\caiji (Windows). Change it to a safe, platform-appropriate directory before running to avoid path errors or accidental writes. - Be mindful of scraping policies and rate: visiting many category pages may be rate-limited or violate a site's Terms of Service — obtain permission if needed and throttle requests. - Because the registry metadata omitted an install spec but code exists, the agent/operator must ensure dependencies and Playwright browsers are installed; otherwise the scripts will fail. If you want me to: (a) generate a safe install/run checklist (commands to run in a sandbox), (b) point out any specific lines of code to change (e.g., default paths, headless flag), or (c) re-audit for any additional concerns, tell me which and I will provide tailored steps.

Type: OpenClaw Skill Name: category-collector Version: 1.0.0 The bundle contains a collection of Node.js scripts (collect.js, collect-ajax.js, etc.) designed to scrape Shopify store category hierarchies using Playwright. The code is well-documented, lacks obfuscation, and performs actions strictly aligned with its stated purpose of web scraping and CSV generation. No evidence of data exfiltration, credential theft, or malicious prompt injection was found in the code or SKILL.md instructions.