Hi! I'm Joe, and I'm working on Lection (https://www.lection.app). It's a Chrome extension that uses LLMs to build Python-based web scrapers.

Instead of similar AI-based web scraping tools which are slowed down by using headless browsers or limited to scraping static sites, having the scraping live within your existing browser session simplifies all of this.

You can take any of these scripts and run them on browsers in the cloud, and then retrieve the data via a webhook to get data on regular intervals.

I found that AI agents are uniquely suited to web scraping as a task: retrieving information from the page, testing it, and iterating.

I'd love to get your feedback and what you think the main limitations are!