sleep

Operational sleep coaching: triage the complaint, run the protocol with numbers, route red flags to a clinician instead of coaching past them. Advise mode only: guide the human, never touch their medication. Diary, trip plans, and preferences persist in ~/Clawic/data/sleep/ (see setup.md on first use, memory-template.md for the file format; created only when the user starts a protocol). If you have data at an old location (~/sleep/ or ~/clawic/sleep/), move it to ~/Clawic/data/sleep/.

When To Use

• User reports trouble falling asleep, 3am waking, early waking, or daytime tiredness.
• Trip planning across time zones: build the light and melatonin schedule before departure.
• Scheduling questions touching sleep: nap timing, caffeine cutoff, workout placement, chronotype.
• Night shifts or rotating schedules: damage-control plan, not adaptation fantasies.
• Nightmares, sleep paralysis, sleepwalking, snoring partner, tracker data, or sleep during pregnancy, new parenthood, menopause, or past 65.
• Not for diagnosing sleep disorders: snoring with gasps, dream enactment, sleep attacks go to a clinician (→ Red Flags). Not for dream journaling (dreams skill).

Quick Reference

Situation	Play
Bad sleep < 3 months, tied to a stressor	Acute: hold wake time, ban naps and early bedtimes, wait it out (→ insomnia.md)
Bad sleep ≥ 3 nights/week for ≥ 3 months	Chronic insomnia (ICSD-3): run CBT-I lite (→ insomnia.md)
Loud snoring + witnessed pauses + sleepy days	Stop coaching, refer for a sleep study (→ Red Flags)
Cannot fall asleep before 2-3am but sleeps fine when free-running	Delayed phase, not insomnia — restriction is the wrong tool (→ circadian.md)
Waking 4-5am done sleeping, often age 60+	Advanced phase vs mood — screen both (→ circadian.md)
Crossing ≥ 3 zones AND ≥ 3 nights there	Adapt: compute Tmin, schedule light by direction (→ jetlag.md)
Crossing < 3 zones OR < 3 nights there	Rule of 3: stay on home time, book meetings in the overlap window
Night or rotating shifts	Anchor sleep + commute light control (→ shiftwork.md)
"Should I nap?"	10-20 min, finished ≥ 8 h before bedtime; never during an insomnia protocol (→ performance.md)
Big day after a bad night, or an unavoidable all-nighter	Damage control: nap math, caffeine timing, no-drive line (→ performance.md)
Tracker score bad, user feels fine	Trust daytime function; stage data is noise (→ trackers.md)
"What supplement helps?"	Melatonin 0.5 mg timed for phase shift; everything else is weak (→ substances.md)
Nightmares, sleep paralysis, sleepwalking, night terrors	Identify by timing and recall, treat or refer (→ parasomnias.md)
Teen, pregnant, new parent, menopause, 65+	Base protocols carry modifiers (→ populations.md)
Room too hot, bright, loud; partner snores; kids or pets in bed	Fix the environment before blaming the sleeper (→ environment.md)
Weekend "catch-up" sleep	Cap wake-time drift at 1 h; 2 h drift = social jet lag and a Monday relapse
Any other sleep complaint	Start the 7-day diary in ~/Clawic/data/sleep/diary.md; no intervention before data

Depth on demand: insomnia.md full CBT-I lite, 3am playbook, relapse plan · jetlag.md direction math, worked trip tables, pre-flight shifting · shiftwork.md anchors, rotation design, first-night survival · circadian.md night owls, larks, DSPS, light therapy · environment.md bedroom, noise, partners · substances.md every cutoff and dose · trackers.md what to read, what to ignore · parasomnias.md nightmares to sleepwalking · populations.md life stages · performance.md naps, debt, all-nighters.

Core Rules

1. Wake Anchor: one fixed wake time ±30 min, 7 days/week; bedtime floats with sleepiness. Highest-leverage single change; check that weekend wake stays within 1 h of weekday wake.
2. Judge sleep by daytime function, not hours. Adult range is 7-9 h (consensus guidelines), not a universal 8: alert on 6.5 h = that user's number; sleepy in meetings after 8 h = a problem despite the hours.
3. Triage before advice: every complaint passes the Red Flags table first. Hygiene tips given to an apnea case cost a year of misdirection.
4. Stimulus control (Bootzin): awake ~20 min by feel (no clock-checking), leave the bed, dim light, boring analog activity, return only when sleepy. Best-evidenced single insomnia technique.
5. Effort inverts in sleep: "try to sleep more" always backfires. Prescribe the opposite: later bedtime, restricted window, worry scheduled earlier in the evening.
6. Time substances by half-life, not by feel: caffeine last dose ≥ 8 h before bed, alcohol last drink ≥ 3 h, melatonin 0.5 mg taken 5 h before target bedtime when the goal is shifting the clock.
7. Light steers the clock and direction depends on timing: light after Tmin advances the clock, before Tmin delays it. Tmin = habitual wake minus 2.5 h (wake 07:00 → Tmin 04:30). Backwards application makes jet lag worse.
8. One intervention per week, measured against the diary. Stacked changes make results unattributable; the diary is ground truth, not memory of the night.

Red Flags

Signal	Suspicion	Action
Loud snoring + witnessed breathing pauses or gasp-awakenings + daytime sleepiness	Obstructive sleep apnea	Refer for a sleep study before any protocol
Dozing while driving, or sleep intruding mid-conversation	Severe sleepiness (Epworth-range > 10)	Refer promptly; advise against driving drowsy now
Acting out dreams: punching, kicking, leaping, mostly age 50+	REM behavior disorder	Neurologist referral, not urgent but not optional
Evening leg discomfort with urge to move, relieved by movement	Restless legs	Clinician; low ferritin is the common driver
Sudden sleep attacks, knees buckling with laughter	Narcolepsy/cataplexy	Sleep specialist
Insomnia + hopeless 3am thoughts, mood collapse	Depression presenting as insomnia	Treat mood as primary; escalate per user's care setup
New snoring or gasping in pregnancy, morning headaches, rising blood pressure	Pregnancy apnea / preeclampsia risk	Prompt obstetric review, not sleep coaching
Night waking driven by pain, reflux, or breathlessness	Medical driver wearing an insomnia mask	Treat the driver first; sleep protocols wait

Anything in this table suspends the protocols in this skill: route to a clinician.

Output Gates

• Did this complaint pass the Red Flags table before any protocol advice?
• Does every melatonin mention carry both dose and clock time (0.5 mg, 5 h before target bedtime for phase shifts)?
• Is prescribed TIB clamped to the floor in insomnia.md and the bedtime phrased as "not before"?
• Are jet lag light windows derived from Tmin converted to destination clock, not from local sunrise?
• Am I prescribing exactly one new intervention this week, with the diary as the measure?
• If the user is a teen, pregnant, postpartum, menopausal, 65+, or on shifts, did I apply the modifiers in populations.md / shiftwork.md?

Configuration

User-dependent variables. Defaults apply until the user states a preference; store them in ~/Clawic/data/sleep/config.yaml.

Variable	Type	Default	Effect
wake_anchor	time (HH:MM)	none	Seeds every derivation: Tmin = wake_anchor − 2.5 h, earliest bedtime, nap cutoff; unset → derive from a 7-day diary
time_format	12h | 24h	24h	Formats every schedule, worked example, and trip-plan table
units	metric | imperial	metric	Bedroom temperature guidance (16-19 °C vs 60-67 °F) and any other physical figure
tracker	text (device name)	none	Tailors trackers.md guidance to the metrics that device reports; none → coach from the diary only

Preference areas to record as the user reveals them:

• schedule — work pattern (office hours, shifts, on-call, freelance), fixed commitments; affects anchor placement and every protocol window
• household — partner schedule, kids, pets, room sharing; affects environment.md plays and stimulus-control feasibility
• substances — what the user actually uses (caffeine dose and timing, alcohol, THC, prescriptions); affects which cutoffs get surfaced first
• risk posture — how aggressively to titrate restriction, how firmly to repeat referrals; affects insomnia.md titration and Red Flags delivery
• reporting — plan format (per-day table vs prose), diary check-in cadence; affects artifacts written to ~/Clawic/data/sleep/

Traps

Trap	Why it fails	Do instead
Sleeping in after a bad night	Dilutes sleep pressure, delays the clock, seeds the next bad night	Same wake time; earlier sleepiness tonight is the repayment
Earlier bedtime to get more sleep	More time awake in bed conditions bed = frustration	Later bedtime until SE ≥ 90%, then extend by 15 min steps
Weekend catch-up ≥ 2 h	Social jet lag (Roenneberg): Sunday-night insomnia, Monday impairment	Cap drift at 1 h; 20-min Saturday nap if needed
Hygiene tips for chronic insomnia	Hygiene alone shows near-zero effect on chronic cases; it is prevention	CBT-I components: restriction + stimulus control
Nightcap for sleep	Onset improves, second half fragments	≥ 3 h alcohol buffer; treat latency with restriction
10 mg melatonin at lights-out for jet lag	Wrong dose and wrong hour; sedation misread as adaptation	0.5 mg, 5 h before target bedtime, eastward only
Morning sunlight on arrival in Europe from the US	Lands before body-clock Tmin, delays the clock, worsens the lag	Sunglasses until converted Tmin, bright light 2-3 h after
Coaching a loud snorer on bedtime routine	Misses apnea; months lost while AHI stays high	Red Flags first, referral before protocol
Treating a night owl teen as an insomniac	Delayed phase + early school start is a clock problem; restriction adds deprivation	Phase-advance protocol (circadian.md), not restriction
"Relax and clear your mind"	Sleep-effort paradox: monitoring for sleep prevents it	Stimulus control; paradoxical intention for high performers
Adjudicating tracker deep-sleep deficits	Stage data is noise at consumer accuracy	Re-anchor on daytime function and the diary

Where Experts Disagree

• Blue light: photobiology shows real melatonin delay; behavioral trials show content arousal dominates in adults. Teens and severe insomniacs get strict screen cutoffs; average adults get engagement rules (no feeds in bed), not amber glasses.
• Napping: performance school prescribes it, insomnia school bans it. Sleeps well → nap freely within the ≥ 8 h cutoff; in protocol → no naps until discharged.
• Melatonin for plain insomnia: trials average ~7 min faster onset; strong effects only for circadian problems. Circadian use yes, nightly-forever use no.
• Chronotype: performance school schedules life around it, clinical school retrains it. Shift the clock only when the phase conflicts with obligations the user cannot move; otherwise move the obligations (circadian.md).

Related Skills

More Clawic skills, get them at https://clawic.com/skills/sleep (install if the user confirms):

• fitness — when the lever is training load, overtraining, or workout timing rather than the night itself
• fasting — when late eating windows or fasting schedules collide with the sleep window
• plan — when the fix is calendar surgery, moving deep work to the user's alert hours instead of fixing sleep
• dreams — dream journaling and pattern exploration; nightmare treatment stays here

Feedback

• If useful, star it: https://clawic.com/skills/sleep
• Latest version: https://clawic.com/skills/sleep

Part of Clawic, the verified skill library. Get this skill: https://clawic.com/skills/sleep.