How Atomus thinks

You read something good. You write it down. A few days later, quietly, a review shows up.

That's Atomus. Three small acts: write, atomize, review. Each one is deliberately lightweight, because the loop only works if every piece is almost free.

Write: notes first

The fastest way to keep something is to write it where you already take notes. In Atomus that's a plain Markdown note: your Inbox for quick things, or the note where the idea belongs.

No decks. No card editor. No special app to switch to. Your notes stay notes, readable in any editor.

Atomize: one line is enough

An atom is a line in your note with a question and an answer, like term == definition, or a sentence with a {{gap}}. Writing it takes a few seconds, and it lives next to the context it came from.

If you'd rather not write the atoms yourself, AI can draft them. Put the cursor on a line and press ⌥ A, and Atomus writes atoms for it using the AI provider you chose. You keep, edit, or delete what it writes. AI is optional; everything else works without it.

Review: come to me, not the other way around

Reviews are the hard part of spaced repetition. Everyone knows the intervals matter. Nobody wants to open an app every day to honor them.

Atomus inverts this. The app comes to you. An atom drops from your notch while you're working in another window, you answer with a key or two, and it's gone. You just kept a memory alive for another few days.

The loop is the product

Writing is cheap. Atomizing takes seconds. Review is ambient. Each piece is small on its own. Together they form a loop that quietly raises the floor of what you remember.

Write what matters, mark what you want to keep, and let the review come to you.