Web Scraper

Web Scraper skill specializes in turning unstructured web pages into clean, structured datasets. Unlike Browser Relay which provides full browser control, Web Scraper is optimized for high-volume data extraction — it's faster, more efficient, and purpose-built for turning page content into JSON, CSV, or spreadsheet-ready data.

The skill uses CSS selectors and XPath expressions generated automatically by Claude based on your description of the data you want. Describe in plain English — 'get all product names, prices, and ratings from this category page' — and the skill figures out the appropriate selectors, handles pagination automatically, and returns a clean array of structured objects.

Supported extraction patterns: e-commerce product listings with price and availability, business directory entries with contact details, real estate listings with price and specs, job postings with salary and requirements, and tabular data from HTML tables. The skill handles common anti-patterns like lazy-loaded content, infinite scroll (with configurable depth), and dynamic class names that change on each page load.

Rate limiting and politeness delays are built in to avoid triggering anti-bot measures and to respect server resources. For JavaScript-heavy sites that require actual browser execution, the skill transparently falls back to Browser Relay mode. Export formats include JSON, CSV, Markdown table, and direct insertion into Google Sheets via the Sheets integration. Marketing researchers, recruiters, and data analysts use this skill daily to build datasets that would otherwise require expensive third-party data providers.

Installation

clawhub install web-scraper

Install: clawhub install web-scraper