Turning one web page into a study space

Paste a link and Novus Learn fetches that page, extracts its readable article text, and runs the same pipeline it runs over a Wikipedia revision: stable sections, source blocks, direct claims, concepts and labelled relationships. The result is a local project, stored in this browser, with every claim one click from the passage it came from.

A recipe, a news report, a tutorial and a reference article carry their meaning in different structures, so the tool classifies the page first and builds the study space that fits it. A page with a numbered procedure gets a how-to breakdown; a page that is mostly prose gets a section breakdown. That is why two URLs of similar length can produce noticeably different study material.

• It does not write anything. Every sentence in the result is a sentence from the page.
• It cannot read a page that requires a sign-in, or one whose publisher blocks automated fetching.
• It does not keep a copy on a server. The project lives in this browser only.

The usual causes are a paywall, an anti-automation check, or a page whose content is assembled entirely by scripts after load. If the article exists somewhere else without those obstacles, try that copy. If you already have the document, the upload route parses it locally and avoids the fetch entirely.