Why URL conversion beats the default web clipper
Web Clipper (and similar browser extensions) ships per-site templates and falls back to generic readability heuristics. The output is usable but variable: cookie banners sometimes survive, ad markers sometimes get clipped as if they were content, footnotes sometimes flatten into prose. A dedicated URL-to-Markdown converter applies consistent stripping across sites, so your vault doesn't accumulate variation that comes from the clipper rather than from the source.
The "literature note from URL" workflow
One Markdown note per source, generated from the URL, with YAML front matter capturing metadata: title, author, url, clipped, tags. The body is the converted content. Then create permanent notes that [[Sources/SmithBlog2026]]-link back, and your graph turns into a real reading-and-thinking system.