Seppi's crawler reading a real website, live: robots.txt first, then the page, its structure as markdown, its metadata and links, the public posts of WordPress sites, and a short walk of the site. This is how Seppi reads the page you give its AI.