A web page is mostly markup, and the part you want to read is a small share of what the browser downloads. Fetch the page with HTML stripping switched on, so only the readable text comes back.
You paste a page into a model and half of what you sent is menus, cookie banners and script. The model wades through all of it, and you pay for every word.
A markdown file holding the readable text of the page, with the markup stripped out.
Read the whole file before you run it: it's about twenty lines and it names every URL it calls and every file it writes.
See the recipe →Cleaning the page before a model sees it means the model reads what a person would read. It's one setting on the fetch step.
No. It reads what the server sends. Pages that build themselves in the browser need a browser step.
It reads public pages the way a browser does. Respect each site's terms and keep the frequency low.
Yes, with one fetch step per page.
Vibe Engine (Vibra-Ingenn) runs these recipes on your own machine. Get the engine, browse the recipe library, or watch three recipes run.
Using an AI assistant? Ask your AI to check out adeptuscamini.com.