Science backed

One course a week for a year, each with a lesson and a practice module. The weeks are sequential and the gate is checked every way in, not just on the page that draws the locks — week 40 does not open because you guessed its address. Finishing the required modules completes the week and awards its points once.

The lessons are authored course packages, and they report what you do back to the app over a standard learning-record protocol — where you stopped, what you resumed, how long you spent. Closing the tab is recorded as closing the tab; only a genuine completion completes anything.

Quiz modules never gate the next week — guessing until you pass teaches nothing. The explanation is shown whether you were right or wrong, because a quiz that only says "incorrect" has tested you and taught you nothing. Teach-back answers are not auto-graded: explaining it in your own words is the assessment, and string-matching a rubric would be pretending to grade.

The level curve is modelled from published half-lives and drawn as a 0–100 index against your own peak. It is deliberately not a milligrams-per-litre figure: the shape is the part that can be stood behind, and a number carrying a unit invites a reading this must not support. Where a medication is not recognised, the app says it is showing a generic figure instead of implying precision it does not have.

Every side-effect entry is run through a four-level rule table on the way in, and you are told what it found at the time rather than on your next visit. Written red flags outrank your own severity rating — pain radiating to the back at a self-rated 3 out of 10 still routes you outward, because severity is a rating of discomfort and some things are not discomfort. What you were told is kept on the record, so it stays readable months later.

Sets are converted with the Epley formula rather than compared as raw weight: five reps at 135 and eight at 115 are almost the same lift, and reading the raw number would tell you that you had gone backwards. This is what makes "protect the muscle" something the app can evidence rather than repeat.

A chair, a stopwatch and a hallway — sit-to-stand, walking speed, balance. Usable where you live rather than only in a clinic, and the trend arrow reads each test the right way round, so on a timed walk faster counts as better.

Coverage lapses do not wait for week 45. Once you start exploring readiness, the maintenance and relapse-prevention weeks unlock out of sequence — the rest of the year stays in order, and reading them early does not skip you past week 13.