What output formats are supported?

The skill defaults to agent-browser bash commands for use within the terminal, but can generate standalone Playwright JavaScript files using the --standalone flag.

What makes Actionbook Scraper different from standard AI scrapers?

It uses a verified selector engine and mandatory two-part verification to ensure data is accurate and not just placeholder text like 'Loading...' or 'Click to expand'.

Can I use this for JavaScript-heavy or lazy-loaded websites?

Yes, the skill includes built-in logic for handling network idle states, scrolling, and waiting for dynamic content to render completely.

How do I request support for a site not currently in the database?

You can use the /actionbook-scraper:request-website command to trigger an analysis and submission for a new URL to be indexed.

Does it handle scraper errors automatically?

Yes, it follows a Generate-Verify-Fix workflow, attempting to resolve syntax, selector, or data mapping errors up to three times before presenting the final code.

Actionbook Scraper

Name: Actionbook Scraper
Author: actionbook

byactionbook

•

1,417

•

网络抓取与数据收集

Generates and auto-verifies resilient web scraping scripts using a database of pre-validated selectors.

The Actionbook Scraper skill streamlines web data extraction by leveraging the Actionbook browser engine's verified selector database. Unlike standard scrapers, it performs a mandatory two-part verification process: ensuring the script executes without errors and validating that the extracted data is accurate, non-placeholder content. It supports both agent-browser bash commands for immediate use and standalone Playwright scripts for external integration, automatically handling complex UI patterns like lazy loading and hidden elements.

主要功能

01Two-Part Auto-Verification: Validates both script execution and data integrity before final output.

02Dynamic Content Handling: Built-in support for scrolling, pagination, and expanding card-based layouts.

03Verified Selector Integration: Uses Actionbook’s resilient selectors to minimize breakage from site updates.

041,417 GitHub stars

05Smart Error Recovery: Automatically detects and fixes selector timeouts, UI text interference, and empty fields.

06Dual-Mode Generation: Creates agent-browser bash commands or standalone Playwright JavaScript code.

使用场景

01Rapidly extracting structured lead or product data from complex, dynamic web applications.

02Building automated data collection pipelines that require high resilience against CSS changes.

03Generating production-ready Playwright scripts for standalone web monitoring and research tasks.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add actionbook/actionbook actionbook-scraper

For use in Claude.ai and ChatGPT

Download Skill