Point, click, and extract. A visual toolkit for building Crawl4AI extraction schemas and automation scripts directly from the pages you're browsing — no code required to get started.
Click on a repeating container (a product card, article block, list item) and then click the individual fields inside it — title, price, description. The extension builds a data extraction schema live, as you click.
Select any element on a page and export it as clean, structured markdown — great for pulling readable content out of articles, docs, and blog posts without wrestling with raw HTML.
Record your browser actions — clicks, typing, navigation — and turn them into an automation script you can replay with Crawl4AI.
Container and field selection with live visual feedback, so you always know what's selected and what mode you're in.
Every schema you build exports as ready-to-run Python code that plugs straight into a Crawl4AI project.
Consistent with Crawl4AI's own design language — built with Manifest V3 for security and performance.
crawl4ai-assistant.zip above to a folder on your computer.chrome://extensions/.crawl4ai-assistant folder (the one containing manifest.json).pip install crawl4ai) and have an LLM API key configured — the exported script
uses Crawl4AI's LLM integration to generate and test the schema.
The packaged extension (v1.3.0) — unzip and load into Chrome.
View the original README with full feature and troubleshooting notes.
View the extension manifest (Manifest V3, permissions, entry points).