if you have a website as a knowledge source, do you have to add each page on that domain separately or is there a way to scrape and follow all links on a domain?