Settle

Three physiological sighs - two in through the nose, the second a top-up, one long out - or four rounds of four in and six out. Then one tap: wired, flat, okay, foggy. Forty seconds.

Parameters

Before every drill. A one-tap skip is always available and the skip is logged without penalty.

Scored

Nothing is scored. The state label is the most useful single piece of data the steering model receives.

Why this shape

Be clear about what this is not. A fifteen-minute single mindfulness-breathing session did not improve inhibitory control or flexibility, and one brief resonance-breathing session did not change sustained-attention or interference scores in anxious adults. But three weeks of daily breath-awareness practice did improve inhibitory control and its brain markers. So the pause is not a performance enhancer taken before a test - the pause is the rep, and the drill is the test. If you want an immediate pre-task boost, three to five minutes of movement is better supported than three breaths.

An expanding and contracting ring with an optional low tone, at about six breaths a minute. On "wired" days, cyclic sighing instead. Four minutes.

Parameters

Ratios 5/5 or 4/6; rate limits 4.5 to 7 breaths a minute; nose breathing, belly, not deep or forced. A "feeling dizzy? breathe lighter" nudge appears if you report lightheadedness. Five to ten minutes a day, most days, for four to eight weeks.

Scored

Resting breathing rate measured from the pacer sessions, over weeks. That is the fair, non-medical number to show improving.

Why this shape

Slow-paced breathing at roughly resonance is the best-supported form of any biofeedback: 0.81 pre–post and 0.83 against control across 24 studies in adults. A 2025 review of remote delivery found medium effects and, usefully for a teen product, that under twenty minutes a day worked better than more - more practice was not better. The individual resonance rate sits between 4.5 and 7 and drifts between sessions in about two-thirds of people, so it is re-measured rather than fixed forever. Equal in-and-out at 5.5 raised variability as well as any variant.

A simple game that gets harder as your heart rate rises and easier as it falls. Cool down to keep playing.

Parameters

Requires a chest strap or arm band over Bluetooth. Scheduled only when the sensor has reported in during the last hour; a strap-off day gets the breathing session instead, with no penalty.

Scored

Minutes in the calm band. Never a numeric variability score unless a beat-accurate sensor produced it, and never as anything but your own trend.

Why this shape

This is the design of the cardiac biofeedback games that have reduced externalising behaviour in proof-of-concept trials, confirmed in a 2025 systematic review, with the largest effects in clinical groups. In children and adolescents the biofeedback base overall is medium-to-large but built on small and often uncontrolled studies, which is why the grade stops at moderate.

Finger on the camera, or a strap. Tones play in or out of time with your heartbeat; say which; get told.

Parameters

Six sessions of about twenty minutes, two tasks per session. Seated only, and the brief jumping step that amplifies the signal is skipped entirely for anyone who reports dizziness on standing.

Scored

Detection sensitivity and confidence calibration.

Why this shape

The ADIE trial (n = 121, six sessions, against an active tone-of-voice control) reduced trait anxiety at three months - 31% "recovered" against 16% - and tracking accuracy rose from 0.48 to 0.76. It is the most direct exercise for the body-sensing half of the profile. Two cautions are built in. Heartbeat counting is a flawed measure, dominated by undercounting and by knowing your own resting rate, so the discrimination format is used instead. And people with orthostatic symptoms underestimate heartbeats and make larger errors under sympathetic activation, and may read body signals as threatening - so onboarding asks about dizziness on standing and routes those users to seated versions, and nothing ever frames a fast heart as a fault.

Thirty seconds a day: when you woke, roughly how you slept, minutes of movement, water.

Parameters

A fixed wake-time goal that you choose, a seven-day regularity score, and a wind-down cue at a time you set that turns into the breathing exercise.

Scored

Regularity - the spread of your wake times - rather than hours. No sleep-debt figure and no alarming numbers.

Why this shape

A 2025 meta-analysis of digital sleep interventions for adolescents found small-to-moderate improvements, with the active ingredients being self-monitoring, goal setting, stimulus control and a consistent wake time - not generic hygiene tips. Hydration gets a tick and no target: school water trials reliably raise hydration and mostly do not show cognitive gains, and for anyone on a clinician's plan, fluid and salt belong to the clinician.

Twenty or more logged minutes of movement with rules - sport, dance, martial arts - is a completed session. The app suggests a drill within the following hour.

Parameters

Three times a week. The drill suggestion sits inside the roughly 30–60 minute window where the acute effect is real.

Scored

Logged, and credited in full.

Why this shape

Both single bouts and regular programmes improve inhibitory control and working memory in young people, with sessions over twenty minutes showing the clearest gains, and open-skill activities beating repetitive aerobic work for inhibition. Crediting movement in full is also the release valve that keeps a daily routine from becoming a cage - which, given the engagement evidence on page three, is not a soft consideration.