E-Commerce Price Monitor

Problem it solves

E-commerce price scraping is often blocked by dynamic page rendering: titles, prices, and store names may live in different selectors, while Taobao and Tmall can hide or omit fields due to login state and anti-scraping controls. This skill targets product pages on jd.com, taobao.com, and tmall.com, turning scattered page data into a clean structured result.

How it works

The skill relies on OpenClaw's built-in browser tool instead of requiring Playwright as an extra dependency:

• Platform detection: identify JD, Taobao, or Tmall from the domain.
• Page load and rendering wait: open the product page, allow dynamic content to settle, then read the snapshot.
• Selector-based extraction: for example, JD pages may use .sku-name, h1 for title and .p-price, .price for price; Taobao and Tmall pages may rely more on .tb-detail-hd h1 and .tb-rmb-num.
• Structured output: return JSON and summarize the product name, current price, store name, and stock status in plain language.

For batch runs, it processes URLs one by one with pauses between requests to reduce anti-scraping risk.

Boundaries and notes

• JD pages are usually easier to scrape; Taobao and Tmall are stricter, so missing title or price may require manual copying or another browser automation setup.
• Full prices may depend on login state, so users should be prompted to retry after logging in.
• Price display normalizes values by removing the ¥ symbol; screenshot OCR is only a fallback and is not recommended as the main approach.