Spaced repetition that stays on your device

A review schedule built around what you remember

A flashcard deck is useful on its first pass, but the real benefit arrives later: seeing a card near the point when its answer might otherwise fade. Novus Learn gives every topic and project a small, deterministic spaced-repetition scheduler. Reveal an answer, judge your recall with Again, Hard, Good, or Easy, and the card receives a new due time. The next session puts unseen and due material ahead of cards that are already scheduled for later.

The routine is deliberately quieter than a streak-driven study app. There is no account, daily target, leaderboard, notification, or server deciding when to pull you back. The schedule exists to reduce needless rereading, not to manufacture engagement. Open the deck when you want to study and the due count tells you what is ready now.

That review layer sits on top of the same source-grounded cards used everywhere else in Novus Learn. Scheduling changes when you see an answer; it does not change the answer, rewrite the source, or make an unsupported card safer through repetition.

The four ratings have concrete jobs

After you reveal a card, choose the rating that describes the recall you just experienced. Again means the answer did not arrive and brings the card back within the same study period. Hard moves it forward cautiously. Good follows the normal review path. Easy gives a stronger first interval and, on mature cards, lets the interval grow faster. These are inputs to a fixed local calculation, not an opaque score inferred from your behaviour.

The first successful reviews use short, predictable steps. Later intervals grow from the card's previous interval and its ease value, with safeguards that keep the schedule within practical bounds. An Again response records a lapse and lowers the ease; an Easy response raises it. You can therefore see the same deck gradually separate into cards that need attention and cards that can safely wait.

• Again reschedules the card for roughly ten minutes and records a lapse.
• Hard advances carefully, so fragile recall returns sooner than a comfortable answer.
• Good uses the standard progression for a remembered answer.
• Easy starts further out and increases the growth factor for later reviews.
• Intervals are deterministic and capped; there is no remote model or hidden personal profile.

What appears first in a session

When a session begins, Novus Learn orders the deck in three groups. Cards you have never rated come first, so new material cannot disappear behind a long queue. Due cards follow, oldest due first. Cards scheduled in the future remain available for browsing, ordered by the soonest due date, but the interface makes it clear that they have already been reviewed and scheduled ahead.

This ordering keeps the deck usable even when you do not follow a perfect daily routine. Miss several days and the overdue material naturally rises. Return early and you can still browse the whole deck without pretending every card is due. Arrow controls let you move freely, while rating a revealed card records the result and advances the session.

The due count above the card is a snapshot of the current local schedule. It is not a universal measure of mastery, and it cannot know whether you guessed or misunderstood the context. The four buttons are a compact way for you to supply that judgment yourself.

The evidence travels with the card

Spaced repetition is only as reliable as the material being repeated. Every Novus Learn flashcard is admitted from claims and concepts found in the open source. Definition cards use defining evidence; cloze and date cards remove a real term from a supported statement; relationship cards depend on a relationship the project can trace. If the source is too sparse to support safe cards, the deck shows an honest empty state instead of filling the gap with plausible trivia.

Reveal a card and its source evidence appears beside the answer. A direct source answer is quoted from the retrieved material; a source-derived answer is assembled from labelled structure and stays marked as such. The evidence link opens the exact source section so you can check wording and context before you rehearse it again.

That check matters most when a short answer feels cleaner than the underlying passage. Retrieval practice makes an answer easier to recall, but it does not make the answer more complete. Keeping the supporting sentence one click away lets you correct the card's place in your understanding before the schedule reinforces it.

Your schedule lives in this browser

Card schedules are saved per project in the browser's IndexedDB alongside the rest of local study progress. The record contains due times, intervals, repetitions, lapses, ease, and the latest rating for each card. Saving a review does not send an event to a Novus Learn account because there is no account and no cloud progress service in this release.

The privacy boundary has a practical consequence: the schedule does not automatically appear in another browser, profile, or device. Private browsing, storage eviction, or clearing site data can remove it. Reset local progress deliberately clears the project's schedule and starts the deck as new. If browser storage is unavailable, the deck still works for the current visit but explains that progress will not persist.

Exports preserve the learning material rather than pretending to be schedule sync. CSV gives you a deck suited to Anki or Quizlet, Markdown creates a readable answer key with source evidence, and the Novus study snapshot keeps the revision-pinned cards and attribution. Your Novus review history remains local unless you separately establish a routine in the tool receiving the export.

A simple routine that does not need a streak

Begin by scanning the summary or source so the deck is testing an idea you have actually encountered. On each card, try to say the answer before revealing it. Rate the recall you had, not the confidence you wish you had. If an answer seems ambiguous, open its evidence immediately; a careful source check is more valuable than preserving the pace of the session.

When the session ends, stop. Cards marked Hard or Again will return sooner, while comfortable cards move out. The next time you open the same project, the local schedule rebuilds the order. A short due queue reviewed honestly is more useful than repeatedly cycling the entire deck until every answer feels familiar.

Spaced repetition cannot decide what deserves to be learned, and it cannot replace working problems, writing, teaching, or reading across multiple sources. It does one narrower job well: it turns an evidence-backed deck into a changing review queue, then leaves the rhythm and the data in your hands.