Questions and honest answers.

Yes. The calendar plans one session a day so the schedule stays honest about how much work you signed up for, but if you finish it and want to keep going, you can ask for another. It is not a repeat of the one you just did: the extra session is built from your state as it stands at that moment, so a point you just answered correctly has already left the review queue and one you just missed is first in line. There is no cap — what stops pointless repetition is the selection, not a daily limit.

Streaks reward showing up, not learning. They also create rebound effects when broken, so a single missed day can demotivate a learner who was actually doing fine. If you want a calendar of when you used the app, that is honest data; a streak counter framed as a moral score is not.

What you remember decays over time. A simple mastery percentage hides that decay and lies about progress. Retention is the estimated probability that you would recall a learning point right now. It drops naturally between sessions and goes back up when you review. The schedule reads this directly to decide what to surface next.

FSRS (Free Spaced Repetition Scheduler) is the algorithm Ignia uses to decide when to review what. It models the forgetting curve for each learning point separately and picks review dates that maximize retention for minimum time. Same family as the Anki algorithm, but newer and more accurate.

The app is free to download and use with your own API key or a local model. A Course Pack covers managed AI for one course, billed up front. Top-ups are available when a course runs short.

Yes, completely. Bring your own API key from any supported provider, or run Ollama locally. The app charges nothing in either case.

Nothing breaks. Every deterministic exercise still works offline. Only open-ended writing and code-review exercises (the AI-judged ones) need credits to be graded. You can either top up your pack or switch the skill to BYOK or a local model.

All your data lives on your device. Once a session is prepared, the practice runs offline. Generation of new lessons and AI grading need a network round-trip. The review queue and the deterministic exercise grading do not.

Locally, in a SQLite database in the app data directory. You can export any course as Markdown or as a portable archive at any time.

Yes, with an account. Sign in with the same email on each device and your courses come with you: the plan, the written course, the exercises, your answer history, your review schedule and the questions you asked the tutor. A session you open on your laptop is the same session on your desktop, rather than being generated again. It happens on its own when you open the app. Sign-in is optional; you can use the app without it.