Rook keeps a live model of what you actually know in SAT math. Every answer — right or wrong — updates it, and the next problem always comes from the frontier: skills the model says you're ready to win.
Behind every problem is a per-skill mastery estimate. The meter you see is the model's real belief — it moves on wrong answers too, because those are evidence.
A curriculum graph decides what's unlockable. You pick your next move from skills you've earned the right to attempt — nothing is skipped, nothing is padded.
No streaks to protect, no fake progress bars. Parents see the same mastery curve the engine sees, with real time controls — and we publish what works.