One loop, seven steps.
Lampira isn't one more note app — it's a system that studies with you. Notes weave themselves into a graph, facts become flashcards, and spaced repetition keeps them sharp. Learn one simple loop and just live inside it. Any subject, any language.
The steps run in order, but it's a loop, not a ladder — every new idea starts a new turn. You don't have to do all seven at once.
The seven steps
-
✍️ Capture
An idea lives for seconds — your job is to catch it, not to format it. Tap + → Quick Note and just write. It saves on your device, offline, instantly.
In the app: the + button, Quick Note, Daily Notes, local-first storage.
-
#️⃣ Tag it
A tag is the first act of sense-making: it turns a pile of notes into a topic. Tap # Tag to add one by hand (autocomplete over the tags you already use), or type
#right in the text. After you save, Lampira shows dashed ghost chips — AI suggestions. One tap accepts, ✕ dismisses.Apple Intelligence suggests 3–5 tags in the note's own language. If it can't (say, Russian), an on-device algorithm always steps in — no language is ever left untagged. The AI proposes; you decide. It never tags silently.
In the app: the tag chip row,
#autocomplete, the auto-tag ladder (Apple Intelligence → algorithm), the “Auto-tag” menu action. -
🔗 Link it
A tag answers “what is this about?”; a link answers “what is this connected to?”. Type
[[to point at another note (or create one on the spot). Link only where the connection is real. Each note lists its backlinks at the bottom.Together, tags and links build your knowledge graph by itself: every note a node, every tag a hub that connects the notes sharing it, every
[[link]]an edge. Nothing to draw by hand.In the app:
[[autocomplete, backlinks, the Graph tab (note nodes, tag hubs, link edges). -
🃏 Turn into cards
Knowledge you never test, you forget. Wrap the decisive fact in a cloze —
{{c1::within 24 hours}}— and that line becomes a flashcard. Or pick Make flashcards from the AI menu and confirm. Hide the minimum meaningful span, not the whole sentence.In the app: cloze markup, the “Make flashcards” action, inline flashcard blocks.
-
🎓 Study
Not “I'll review someday” but “I clear today's queue.” The Study tab shows what's due; take it to zero. FSRS schedules each card for the moment you're about to forget it. Grade honestly — a miss is a signal, not a failure.
In the app: the Study tab, FSRS scheduling, the daily queue, reminders, streaks.
-
🗓️ Plan
An exam is a date, not a mood. In Exam mode pick the cards and set the date — FSRS paces the repetitions so every card is ready on the day, and tells you honestly when a date is too close. The Calendar keeps lectures, deadlines and daily notes next to your queue — write on the Mac, review on the phone; the plan follows you through your own iCloud.
In the app: Exam mode, the Calendar tab (categories, reminders, daily notes), iCloud sync, iPhone · iPad · Mac.
-
✨ Ask
Your notes are a conversation partner. Tap Ask AI to question them, get quizzed, or pull a summary — answered on your device, with citations back to your notes. Nothing leaves it. Prefer another brain? Connect a local model (Ollama / LM Studio), bring your own API key (Gemini or any OpenAI-compatible provider), or use the MCP connector inside your own Claude, ChatGPT or Grok.
In the app: Ask (chat over your notes), source citations, on-device AI, a local-model fallback.
A walkthrough for every kind of work
The same loop serves very different people. Here's how it lands.
Lectures land in Quick Note; tags like #hematology gather the topic; [[Iron metabolism]] links the pathophysiology; key facts become cloze cards; the daily Study queue spreads 1,500 cards across eight weeks; “Quiz me on anemias” closes the loop.
Capture the gist of a norm or ruling; tag by area and article (#contract); link a statute to the cases that applied it — backlinks reveal the whole line of practice; cloze the elements and exceptions; review keeps wordings sharp for the hearing; Ask surfaces every note on a topic.
Write it down while the context is hot; tag by tech and area (#kubernetes); link patterns like [[Idempotency]] the way real architecture does; cloze commands, flags and boundaries; Study keeps rare-but-critical details from fading; Ask pulls up your own docs.
Catch a word with its context; tag by level and theme (#B2); link word families ([[get]] ↔ [[get over]]); two-sided vocabulary is perfect cloze material; the daily queue is your practice; “Drill me on B2 phrasal verbs,” in your language.
One Quick Note per source (hypothesis, method, result, limit); topic tags stitch sources by problem; links to concepts show where the literature thickens; cloze the key facts and figures; review keeps the evidence base in your head; “Summarize my notes on spaced repetition,” with citations.
Anything new from class goes into Quick Note; subject = tag (#biology); link topics that belong together; cloze the key fact or auto-make cards from a chapter; five minutes of Study a day beats an all-nighter; Ask is your tutor.
Why it works
Run the loop once, and the system studies with you:
- notes become a graph (Capture + Tag + Link),
- facts become cards (Turn into cards),
- repetition keeps them sharp (Study),
- and your notes answer back (Ask).
Study the system once; let the system study you forever.