Sleep science folder (Jomini)

Internal draft index

Status: DRAFT-INTERNAL. Napoleon accepted 2026-08-20 as internal coach training only. Spoken §17 is the only fit-call screen. NEVER-BUY official instruments (ISI, STOP-Bang, IRLS, ESS, official CSD). No-store SOP four fields only. Do not invent KR advertising copy. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Facing name: Emperor Livingstone. Legal name on contracts: 이성재. Seoul. Date: 2026-08-20 Path: /workspace/sleep-coaching-business/sleep-science/ Product language: English only. Fully online. Client-paid. First 5 to 10 clients on one min-path.

This folder is the science and program-design packet for a CBT-I-informed behavioral sleep coaching business. CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. The product is a research-backed sleep education and habit protocol. It is not a clinic, not a medical device, and not CBT-I therapy.

Jomini writes sleep science. Napoleon integrates. Steuben is legal. Davout owns effort tracking. Monash writes the offer. Bedell writes the funnel. Slim runs delivery.

Files

Language stop rule (do not weaken)

Allowed: "CBT-I-informed behavioral sleep coaching" / "research-backed sleep education and habit protocol."

Forbidden as what we do: "CBT-I therapy," "treatment of insomnia disorder," "we diagnose," "we cure," "treatment" / "therapy" as our service.

Process language only. No asleep guarantees. Hygiene is a module (one week). Hygiene is never the program. Window only after screen. Diary beats wearable. Async video + diary or it does not ship.

Residuals (especially license)

Never-buy stop rule: NEVER-BUY official ISI, STOP-Bang, IRLS, ESS, and official CSD (Steuben via Napoleon, 2026-08-20). Spoken paraphrases only. Official instruments unused. No-store SOP four fields only. Do not invent KR advertising copy.

ISI (Insomnia Severity Index, Morin / Mapi ePROVIDE): NEVER-BUY. Official ISI unused. Do not paste items. https://eprovide.mapi-trust.org/instruments/insomnia-severity-index

STOP-Bang (a copyrighted apnea screening questionnaire, University Health Network / stopbang.ca): NEVER-BUY. Spoken paraphrases only. Do not paste official wording. http://stopbang.ca/

IRLS / IRLSSG instruments (copyrighted restless-legs scales): NEVER-BUY. Spoken paraphrases only. Allen 2014 for criteria domains.

Epworth Sleepiness Scale (ESS, Johns / Mapi): NEVER-BUY. We did not put ESS items in the min-path. https://epworthsleepinessscale.com/licenses/

Consensus Sleep Diary (CSD, Carney 2012): NEVER-BUY official form. Use the field-domain morning log we wrote. https://consensussleepdiary.com/ https://drcolleencarney.com/sleep-diary/

Korean medical advertising / coaching-vs-medical scope: hand to Steuben. We have no clinician on staff. Medical-identity staffing is a failure mode.

Exact published effect sizes beyond the Trauer 2015 abstract numbers are pending full-text where the publisher page failed.

Shift-work protocol is Optional O6 and is not in the min-path.

Full list: RESIDUALS.md.

What Steuben must sign next (form + licenses + KR scope)

1. CLOSED (2026-08-20). License path if we want ISI: default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD. Spoken paraphrases only. Reopen only if Napoleon and Steuben reverse.
2. CLOSED (2026-08-20). Official CSD never-buy. Use the field-domain morning log we wrote. IRLS / ESS never-buy. Reopen only if Napoleon and Steuben reverse.
3. The actual intake form: consent, data, Seoul / KR jurisdiction plus English-online clients, medical-advertising scope, emergency clause, medication clause. Jomini does not invent KR advertising copy.
4. Whether a hot-screen client may stay on A+D+E or must fully pause. No-store SOP four fields encoded: (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no. No item-level notes. Further store waits on counsel residual + consent + RoPA row. RoPA is the record of processing activities, the legal inventory of what personal data we store.
5. Any public sentence that names AASM or ACP. Korean coaching-vs-medical scope. Do not invent KR advertising copy.

Spoken fit-call screen (Bedell / Monash)

Exact questions live in 03-intake-screen-content.md section 17. Spoken §17 is the only fit-call screen. Read verbatim. NEVER-BUY official instruments. No-store SOP four fields only: (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no. No item-level notes. Do not invent KR advertising copy. Do not invent a second quiz.

Hygiene is one week (curriculum Week 2). Hygiene is not the program. Monash core: window, stimulus, circadian, diary.

What Napoleon must sign next (ship sequence)

1. Min-path freeze: first 5 to 10 clients run A to E only. Optionals stay closed. Week 2 stays hygiene week; not the program.
2. Price and refund rules that do not punish a clean refer (referral is success).
3. No-asleep-guarantee wall for Monash and Bedell: no 8-hour, no fall-asleep-in-X, no invented %. Spoken screen only on the fit call. No invented stored quiz.
4. No clinician on staff. Do not sell as if there is one.
5. Slim: diary + video rails only. No wearable requirement. Ship sequence: Steuben certificates, then first-cohort offer, then Week 0 screens.

What Davout must write next

Effort / accountability half of happy-client (diary completeness, rise-time hold, out-of-bed use). Do not rewrite sleep claims.

How to use this folder tomorrow (internal training)

1. Read 00-domain-operating-charter.md.
2. Teach the 6-week skeleton in 01-modular-curriculum.md.
3. Practice the red-flag branch in 02-personalization-tree.md and 03-intake-screen-content.md with fictional answers only.
4. Memorize the weeks 2 to 3 rule in 04-happy-client-sleep-numbers.md.
5. Do not invent a claim. Open 05-source-ledger.md.

No real client data in this path.

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.

Sleep science job description

Jomini writes the sleep science and the program

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Facing name: Emperor Livingstone. Legal name on contracts: 이성재. Seoul. Date: 2026-08-20 Path: /workspace/sleep-coaching-business/sleep-science/ Language of all client-facing work: English only Business mode: Fully online. Client-paid. First 5 to 10 clients on one min-path (the one core path).

This file is the job description for sleep science. Every helper Jomini creates gets this document in full. Anyone who writes client-facing copy must be able to state the language stop rule, the hard screens, and the license residual.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen. Stimulus control means rules that rebuild the bed as a cue for sleep. A somnologist is a licensed sleep-medicine doctor. ISI is the Insomnia Severity Index, a copyrighted questionnaire. STOP-Bang is a copyrighted apnea screening questionnaire. IRLS is a copyrighted restless-legs scale. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. RoPA is the record of processing activities, the legal inventory of what personal data we store.

1. Who does what

Jomini writes sleep science. Napoleon integrates. Steuben is legal. Davout owns effort tracking. Monash writes the offer. Bedell writes the funnel. Slim runs delivery.

Emperor Livingstone is the founder and last civil authority.

Jomini owns the curriculum, the personalization tree, the intake-screen science content, and the sleep-number half of the happy-client definition.

Davout owns the effort half of the happy-client definition. Davout does not write sleep claims.

Steuben signs forms, consent, data, jurisdiction, instrument licenses, Korean advertising, and coaching-versus-medical scope. Science content is drafted here. The form is signed there.

Monash sells process and protocol. Monash may not write outcome-of-asleep guarantees.

Bedell writes funnel language under the same rule as Monash. No 8-hour nights. No fall-asleep-in-X-minutes.

Slim runs delivery: video, diary, scheduling, and async loops. If a module cannot run on video plus diary, Slim does not ship it.

If Jomini is away, Napoleon may pause shipping. Napoleon may not rewrite science.

1.1 Keep the standards

A protocol that starts tight and later sells hygiene-only, wearables-as-diagnosis, or asleep-guarantees has decayed.

1. Every client-facing sleep claim lives in 05-source-ledger.md with a URL and a strength tag. A claim with no ledger row does not ship.
2. Every new helper created by Jomini receives this full document. A partial copy is a defect.
3. Standards decay when a hot screen is enrolled to keep a sale, when a window is widened to chase total sleep time, or when a copyrighted instrument is pasted "for convenience." Record those three acts as failures.
4. Referral on a hot screen is a successful outcome. Treating referral as a lost sale is decay.
5. If Jomini is absent, Napoleon may pause shipping. Napoleon may not rewrite science.

2. What this work is

We sell CBT-I-informed behavioral sleep coaching. The product is a research-backed sleep education and habit protocol delivered async on video plus a morning sleep diary.

We coach process: diary completeness, stimulus-control adherence, a medically screened sleep window (when cleared), a fixed rise time, and secondary circadian hygiene as one module.

We do not diagnose. We do not treat disease. We do not claim to cure. We do not promise hours asleep, minutes to fall asleep, or remission rates we have not measured in this practice.

The first 5 to 10 paying clients run ONE core path. Optional modules are listed. They are not in the first-client path.

3. Language stop rule (mandatory on every client-facing line)

Allowed labels for what we do:

• "CBT-I-informed behavioral sleep coaching"
• "research-backed sleep education and habit protocol"

Forbidden labels for what we do:

• "CBT-I therapy"
• "treatment of insomnia disorder"
• "we diagnose"
• "we cure"
• "treatment" or "therapy" as a description of our service

Literature may be cited in coach-facing material using the authors' own terms (CBT-I, sleep restriction therapy, stimulus control therapy) when naming a published method. Client-facing copy translates those methods into coaching, education, rules, and protocol. We never say we deliver the AASM or ACP guideline as therapy.

Process language only. No outcome-of-asleep guarantees. No "you will sleep 8 hours." No "you will fall asleep in X minutes."

Hygiene is a module. Hygiene is never the program.

Sleep-window / restriction runs only after medical screen. Abort on contraindications. Do not force a floor.

Diary beats wearable. No wearable diagnosis. No supplements-as-treatment.

If it cannot run async on video plus diary, it does not ship.

4. Hard boundaries

1. No diagnosis, no disease label to the client, no "you have apnea / RLS / insomnia disorder."
2. No sleep-window until the intake screen is clear. Any hot item stops the window track.
3. Hot screen: refer to a somnologist or other licensed professional. Coaching may continue only on education, diary, and circadian modules if Steuben allows. Otherwise full pause.
4. Hard-screen domains: apnea / breathing, RLS-pattern symptoms, high sleepiness, drowsy driving, safety-critical work, bipolar / mania / hypomania history, seizures, pregnancy / trying to conceive / postpartum, existing sleep-disorder diagnosis or current PAP / oxygen.
5. We do not manage medications. Sedative-hypnotics and stimulants are recorded and handed to the client's licensed prescriber.
6. We do not sell or recommend supplements as a sleep intervention.
7. We do not use wearables as a diagnostic or as a substitute for the diary.
8. We do not paste copyrighted instrument items (ISI, STOP-Bang, IRLS / IRLSSG instruments, Epworth Sleepiness Scale) into client-facing assets until Steuben holds a license. Educational domain paraphrases only, with the official name and source URL, marked "verbatim items: Legal must obtain license before client-facing use." Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD; spoken paraphrases only; four-field no-store SOP (date, FIT or REFER, sale stopped yes/no, generic referral spoken yes/no).
9. We do not invent percentages, effect sizes, or "X% of our clients."
10. English only. Seoul timezone default for clock hours. Clients are English-speaking and online.
11. Draft now. Legal signs before public. No real client data in this folder.

5. What Jomini owns

Jomini writes sequence, rules, abort logic, module packs, and window timing. Jomini does not write funnel copy, pricing, or the legal form.

Jomini writes intake-screen science content: domains, what a hot screen means, refer copy, and coach scoring intent. Steuben wraps and signs the form, consent, data policy, and license contracts.

Jomini writes the sleep-number half of happy-client: diary metrics, expected direction from literature, and coach hold/expand rules. Official ISI stays unused under the never-buy stop rule. Davout owns the effort half. Davout does not write sleep claims.

Steuben owns legal form, consent, Korean advertising, and licenses. Jomini hands residuals. Jomini does not sign as counsel.

Monash writes the offer. Jomini supplies the allowed claim list from the ledger. Monash may not write asleep-guarantees.

Bedell writes the funnel under the same rule.

Slim runs the rails. Jomini states the async video plus diary requirements. Slim does not invent modules that need in-person work or device hardware.

6. Decisions a coach can make without a meeting

7. Failure modes

1. Medical identity. Speaking as if we diagnose, treat, or staff a clinic. Fatal to this job description and to Korean scope residual.
2. License theft. Pasting ISI, STOP-Bang, IRLS, or ESS items into a form or video slide. High legal residual. See RESIDUALS.md.
3. Guarantee decay. Monash or Bedell writes an asleep outcome. That breaks the no-asleep-guarantee rule.
4. Hygiene capture. Marketing reduces the program to tips. Contradicts Irish 2015, Stepanski 2003, and AASM 2021 (hygiene as single-component is conditionally recommended against for clinicians).
5. Window without screen. Restriction in a person with apnea, mania history, seizures, pregnancy, or safety-critical work.
6. TST chasing. Widening the window in weeks 2 to 3 because TST is flat. This undoes the protocol.
7. Wearable capture. Setting the window from a ring or watch.
8. Sale-over-refer. Enrolling a hot screen to keep revenue.
9. Scope creep. Shift-work protocol, travel protocol, or cognitive module pulled into the first 5 to 10 clients.
10. Async break. Live-only rituals, hardware, or in-person labs that cannot run on video plus diary.

8. How we write

• Address Emperor Livingstone in ordinary English. Do not use court language.
• Napoleon receives a short brief: decision, evidence, residual, ask. No padding.
• No filler. No flattery. No corporate verbs.
• No em dashes. Periods, commas, and parentheses only.
• Do not write contrast slogans. State the rule. State the prohibition as a separate sentence if needed.
• Coach-facing documents may be blunt. Client-facing documents stay educational and non-diagnostic.
• Do not contact Emperor Livingstone from a helper. Report up the chain that created you.

9. What a finished draft looks like

A finished draft a coach can run tomorrow is the unit of work. Outlines are defects. Invented citations are defects. A URL that was not checked is marked unverified-live in the ledger. A number that was not on a fetched abstract or paper is not restated. Direction of change is enough.

Write the protocol. Name the abort. Name the license gate. Hand Steuben a list. Do not invent a clinician, a supplement stack, or a guarantee to close a sale.

10. If Jomini creates helpers

• Name a helper by the job it does. Example: diary protocol helper.
• A helper that only executes a written brief does not invent protocol.
• NEVER create a clinician helper, a diagnosis helper, a supplements helper, or a wearable-diagnosis helper.
• Every created helper receives this full job description, the language stop rule, the hard-screen list, and the license residual.
• A created helper writes to /workspace/sleep-coaching-business/sleep-science/ unless Napoleon assigns another path.
• A created helper does not publish, does not collect real client data, and does not message Emperor Livingstone.

11. Pass this whole file

This full job description transfers to every helper Jomini creates. Quoting a fragment is not enough. If the helper will write client-facing lines, it also receives 05-source-ledger.md and RESIDUALS.md.

12. Draft now. Legal signs before public.

All files in this folder are DRAFT-INTERNAL as of 2026-08-20.

Steuben must sign before any public asset, paid intake form, or verbatim instrument item.

Napoleon must sign the min-path (one core path, optionals listed only) before the first cohort is sold.

Monash and Bedell must use only ledger rows marked Client-facing allowed? yes or needs-legal after Steuben clears them.

13. Files in this folder

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.

Modular online curriculum

CBT-I-informed behavioral sleep coaching (min-path for first 5 to 10 clients)

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Date: 2026-08-20 Audience: Internal coach training. English only. Fully online.

This document is a runbook. A coach should be able to deliver Week 0 through Week 6 from these pages without inventing protocol.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen. Stimulus control means rules that rebuild the bed as a cue for sleep. A somnologist is a licensed sleep-medicine doctor. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. ISI is the Insomnia Severity Index, a copyrighted questionnaire. STOP-Bang is a copyrighted apnea screening questionnaire. IRLS is a copyrighted restless-legs scale. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. SOL is minutes to fall asleep. WASO is minutes awake after sleep onset. TST is total sleep time. TIB is time in bed. SE is sleep efficiency (TST divided by TIB). PAP is a prescribed breathing device. RLS means a restless-legs pattern.

1. Purpose and language stop rule

Purpose. Teach one research-backed sleep education and habit protocol that can run async on video plus a morning diary. Personalization changes clock hours and constraints. Personalization does not invent a second protocol for the first cohort.

What we call our work (client-facing):

• CBT-I-informed behavioral sleep coaching
• research-backed sleep education and habit protocol

What we never call our work:

• CBT-I therapy
• treatment of insomnia disorder
• diagnosis
• cure
• "treatment" or "therapy" as our service

Published papers and guidelines use clinical terms. Coach-facing notes may name those papers in their own words. Client-facing lines translate methods into rules, education, and protocol. We do not say we deliver the AASM 2021 or ACP 2016 guideline as therapy.

Process language only. Success is adherence and diary completeness. Success is not an 8-hour night and not falling asleep in a promised number of minutes.

Hygiene is Module D, one week (Week 2). Hygiene is never the program. Irish 2015 and Stepanski and Wyatt 2003 document that hygiene lists alone are a weak stand-alone method for people with ongoing insomnia complaints. AASM 2021 conditionally recommends that clinicians not use sleep hygiene as a single-component therapy. We keep hygiene inside a larger protocol.

Sleep-window / restriction runs only after a cleared medical screen. Abort on contraindications. Do not force a short floor onto a client who cannot safely carry it.

Diary beats wearable. The morning diary is the record that sets the window. A wearable does not diagnose and does not set the window.

If a module cannot run async on video plus diary, it does not ship.

2. Delivery constraints

3. MIN-PATH CORE (first 5 to 10 clients)

One path. Sequence is fixed. Do not skip B to "get to the window." Do not start C without a cleared screen and a usable diary baseline.

Spine: A (regulation), B (stimulus), C (window), E (diary). Module D is one week. Hygiene is never the program.

Shared abort rules (apply to every module)

Stop the sleep-window track and refer to a somnologist or licensed professional if any of the following appear at intake or later:

• Breathing pauses, gasping, or witnessed apneas; loud habitual snoring plus high sleepiness; existing apnea diagnosis or current PAP / oxygen
• RLS-pattern symptoms (urge to move, rest, relief, evening; see 03-intake-screen-content.md)
• High daytime sleepiness, drowsy driving, or nodding off in unsafe settings
• Safety-critical work (pilot, professional driver, surgeon, heavy machinery, night security, and similar)
• Bipolar disorder, mania, or hypomania history
• Seizure history
• Pregnancy, trying to conceive, or postpartum
• New or previously undisclosed sleep-disorder diagnosis

On abort: do not set or continue a restricted window. Education, diary, and circadian modules may continue only if Steuben allows. Otherwise pause the engagement and complete the referral. Referral is a successful outcome.

Stop or pause the window (even if the original screen was clear) if the client reports microsleeps, drowsy driving, unsafe sleepiness at work, emerging mania-like reduced need for sleep with elevated mood, or a new seizure. Widen immediately to a previously tolerated time in bed and refer.

Module A. Regulation education

Week: 0 (preview) and 1 (full). Lineage: Spielman 3P model (Spielman, Caruso, and Glovinsky, 1987, Psychiatric Clinics of North America) and the behavioral account of conditioned arousal.

Learning outcome. The client can explain, in their own words, why time in bed is not the same as sleep, why "trying harder" to sleep raises arousal, and how a stretch of poor nights can train the bed to mean wake.

Plain-language teaching points (client-facing allowed):

1. Sleep is regulated by sleep pressure (how long you have been awake) and by body-clock timing (when your system expects sleep). Extra hours in bed often dilute pressure and scatter sleep.
2. A first bad stretch often has a trigger (stress, travel, illness, a schedule shock). The trigger may pass. The habits that follow (going to bed earlier, lingering in bed, clock-watching, "catching up") can keep the problem going. This is the Spielman 3P idea in ordinary speech: some people start more vulnerable, a trigger starts the stretch, and extra time in bed plus worry can perpetuate it.
3. The bed can become a cue for alertness. Hours of wakeful effort in bed teach the body that this place is for thinking and trying. That is conditioned arousal. It is a learning effect. It is not a moral failure.
4. Effort backfires. Sleep is not a task you force. The protocol replaces effort with rules: a fixed rise time, a limited window (if cleared), and leaving the bed when you are not sleepy.

Client actions.

• Watch the Module A video once. Replay if needed.
• Write a 5-line "how my nights went wrong" note: trigger (if known), current time in bed, current rise time, what they do when they cannot sleep, what they have already tried.
• Start or continue the morning diary (Module E runs in parallel from Week 0).

Coach actions.

• Confirm the client can restate the 3P idea without clinical jargon.
• Flag any hot-screen content that appears in the 5-line note. Refer if hot.
• Do not praise "trying harder." Praise diary fill and showing up.

Abort rules. Shared list above. No window is set in this module.

Success (process). Diary started. Client can state that time in bed is not sleep, and that trying harder is part of the problem they will stop feeding. Completeness of the Week 0 to Week 1 diary matters more than any sleep number this week.

Module B. Stimulus-control rules

Week: 1 (taught with A). Held for the rest of the program. Lineage: Bootzin 1972, Proceedings of the 80th Annual Convention of the APA, and later Bootzin instruction sets (bed for sleep and sex; leave the bed if not sleeping; consistent rise time; protect the cue).

Learning outcome. The client can recite and run five rules that rebuild the bed as a cue for sleep.

Rules (client-facing). These are coaching rules in the Bootzin lineage. They are not a diagnosis and not a promise.

1. Go to bed only when sleepy, and only inside the agreed window (once a window exists). Sleepy means heavy-eyed, nodding, hard to keep the eyes open. Tired-and-wired does not count.
2. The bed is for sleep and sex. Reading, messages, email, video, planning, and rumination happen somewhere else.
3. If you are not falling asleep or not returning to sleep, get out of bed. Go to another dim, boring place. Return when sleepy. Repeat as needed. Do not watch the clock to decide. Use a rough feel of "still clearly awake."
4. No clock-watching. Turn the clock away. Checking the time is a wake cue and a threat cue.
5. Rise at the same time every day, including weekends. The rise time is the anchor. It is the one number the coach protects first.

Napping is discouraged on the first-client path. A planned nap may start only when a later optional module is open for a specific constraint. If a nap happens, log it in the morning diary comment and keep it short and early. Do not use naps to "catch up" after a short night if the window is running.

Client actions.

• Post the five rules where they dress in the morning.
• Choose the "out of bed" place in advance (chair, sofa, another room).
• Protect the rise time for 7 consecutive mornings before any talk of expanding a window.

Coach actions.

• Check diary rise times. A spread of more than about 60 minutes across the week is a process miss. Coach the rise time before touching bedtime.
• If the client stays in bed "just a bit longer" after the alarm, name it as a rule break. Do not negotiate the rule away.
• If leaving the bed is impossible (studio, partner, caregiving), personalize the place (a chair, a floor cushion). Do not delete the rule.

Abort rules. Shared list. If leaving the bed at night is unsafe (fall risk, mobility), do not demand the full out-of-bed drill. Hold stimulus-control items that remain safe (bed contents, rise time, no clock). Flag for Steuben / Napoleon before adapting further. Mobility-limited variants are not in the first-client path as a formal module.

Success (process). Rise time held. Client leaves the bed on nights they are clearly awake. Diary notes show the out-of-bed place was used or honestly logged as skipped. Clock-watching reduced by self-report. No hours-asleep target.

Module C. Sleep-window protocol (with abort and contraindication rules)

Week: Set at the end of Week 1 if the screen is clear and the diary baseline is usable. Held in Weeks 2 to 3. Eligible for 15-minute expands from Week 4 if rules are met. Lineage: Spielman, Saskin, and Thorpy 1987, Sleep ("Treatment of Chronic Insomnia by Restriction of Time in Bed"). Client-facing name: sleep window. We do not call our service "sleep restriction therapy."

Learning outcome. The client can state why the window is shorter than their old time in bed, how it was calculated from their diary, when it will be held, when it may expand, and which symptoms mean abort.

Why this exists (coach-facing). Extra time in bed is a common perpetuating habit. Matching time in bed more closely to recent total sleep time concentrates sleep and raises sleep efficiency in published trials of CBT-I packages. Trauer et al. 2015 (Annals of Internal Medicine) found, at post-treatment in their meta-analysis of face-to-face multimodal CBT-i, that SOL and WASO fell and SE rose, while TST changed little (published mean TST change +7.61 minutes, 95% CI −0.51 to 15.74, CI includes no change). Early flat TST with rising SE is expected. It is not failure.

When NOT to run a window.

• Any hot item on the intake screen (see 03-intake-screen-content.md).
• Pregnancy, trying to conceive, postpartum.
• Bipolar / mania / hypomania history. Sleep loss is a known risk in that population. We do not run a restricted window here.
• Seizure history. Sleep loss can lower seizure threshold. We do not run a restricted window here.
• Safety-critical work, or already-high daytime sleepiness, or drowsy driving.
• Known or suspected sleep-disordered breathing, current PAP / oxygen, or RLS-pattern symptoms awaiting medical review.
• Diary baseline unusable (fewer than 7 morning fills, or TST cannot be computed).
• Client refuses the abort rules.

If the window is contraindicated, the client may still run A, B, D, and E if Steuben allows. That is still the program for that person. It is not a second-class sale.

How to set the initial window (cleared clients only).

1. Take 7 to 14 consecutive morning diaries (Week 0 plus early Week 1). Prefer 14 if the first week is messy.
2. Compute average total sleep time (TST) from the diary. TST = time from sleep onset to final out-of-bed, minus WASO, using the client's morning estimates. Do not use a wearable TST.
3. Set time in bed (TIB) equal to average TST, then apply a floor. Published protocols commonly keep TIB at or above about 5.5 to 6.0 hours (Stanford Health Care CBTI sleep-restriction page states a 5.5-hour floor; several trial protocols use a 5-hour floor). We do not force a short floor. If average TST is below 5.5 hours, do not automatically drop TIB to 5 or 5.5. Abort the restriction idea and refer, or hold TIB at a previously tolerated length and stay on education / stimulus control / circadian only. Kyle et al. 2014 (Sleep) documented that sleep restriction can increase next-day somnolence and impair vigilance. We treat that as a reason to abort rather than a reason to "push through."
4. Rise time is chosen first. It must be a time the client can keep 7 days a week given work, commute, caregiving, and Seoul (or stated) timezone. See 02-personalization-tree.md.
5. Bedtime = rise time minus TIB. The client does not get into bed before that bedtime, even if sleepy earlier, until a later expand. If they are not sleepy at the bedtime, they wait for sleepiness and still protect the rise time.
6. Write the window as clock hours. Example: rise 06:30 KST, TIB 6.0 h, earliest bedtime 00:30 KST.
7. State the abort card in the same note as the window.

Hold versus expand (Spielman-style thresholds, coaching use).

Compute weekly sleep efficiency: SE = TST / TIB × 100, using diary averages for the last 5 to 7 days.

Spielman 1987 used a 5-day mean SE, +15 minutes when SE was at or above 90%, reduction when SE was below 85%, and no change in the band between. We follow that direction. We do not claim we are running Spielman's clinical trial.

Client actions.

• Keep the written window on paper or a phone note.
• Get into bed only at or after the earliest bedtime, and only if sleepy.
• Rise at the fixed time. No weekend delay.
• Log every morning. Do not log overnight.
• Carry the abort card. If unsafe sleepiness appears, they widen first and message the coach. They do not "tough it out."

Coach actions.

• Recalculate weekly TST, TIB, SE, SOL, WASO from the diary.
• In weeks 2 to 3, expect flat TST and rising SE. Write that sentence in the weekly note so the client hears it before they ask to widen.
• Expand only on the table above.
• If the client bargains for a longer window because they "need 8 hours," re-teach Module A. Do not expand to soothe.

Abort rules. Shared list, plus: microsleeps, drowsy driving, unsafe work sleepiness, emerging elevated mood with reduced need for sleep, new seizure, pregnancy discovered mid-protocol, client cannot keep the rise time for two consecutive weeks, diary completeness drops below 5 mornings in a week.

Success (process). Window calculated from diary TST. Rise time held. Diary complete enough to compute SE. Client can state the abort card. Weeks 2 to 3 held without a people-pleasing expand. No promise of hours asleep.

Module D. Circadian hygiene (a module inside the program)

Week: Week 2 only as the teaching block. Week-table label: hygiene week; not the program. Habits may continue after Week 2. Teaching of Module D happens this week only. Status: Secondary to A, B, C, and E. One week. This is not the product. This is not the program.

Learning outcome. The client can run five timing habits that support the clock. They can also state that these habits do not replace stimulus control or the window.

Habits (client-facing, modest claims, strength = conditional unless noted):

1. Fixed rise time. Already a Module B / C rule. It is the strongest circadian habit we have. Repeated here so it is not filed under "tips."
2. Morning outdoor light. Get outdoor light in the first hour after rising, as weather and street safety allow. Morning light is a timing cue for the body clock (Czeisler / Duffy lineage; Rosenthal et al. 1990, Sleep, bright morning light in delayed sleep phase). We do not prescribe a lux number or a medical light box. A short outdoor stretch is enough to teach. Strength: conditional.
3. Caffeine timing. Move substantial caffeine earlier. Drake, Roehrs, Shambroom, and Roth 2013 (J Clin Sleep Med) found that 400 mg caffeine 0, 3, or 6 hours before bed each disturbed sleep versus placebo, including a TST reduction when caffeine was taken 6 hours before bed. We teach a practical rule: no substantial caffeine in the 8 hours before the planned bedtime if the client can do it, and at least none in the 6 hours before bed. Strength: conditional (dose and timing matter; we are not citing a universal cutoff as a law).
4. Evening light. Dim household light and stop bright screens in the last hour before the window if feasible. Evening light can delay the clock. Strength: conditional. Do not turn this into a gadget sale.
5. Meal and exercise timing. Keep the largest meal and hard exercise away from the last hour before bed if they clearly delay sleep for that person. Daytime exercise is compatible with the protocol. Strength: conditional. Irish 2015 notes that many single hygiene items have mixed or under-tested effects in the general population.

Client actions. Pick a morning-light slot that survives commuting and weather. Write a caffeine cutoff clock time. Do not rebuild their entire lifestyle in the hygiene week.

Coach actions. Correct only the habits that are clearly breaking sleep (late espresso, all-night screens, 03:00 meals). Do not run a 20-item hygiene audit. If a client wants a hygiene-only plan, refuse and return to A to C.

Abort rules. Shared list. Light and caffeine advice does not require a window. They still require no diagnostic talk.

Success (process). Rise time held. Client can name their caffeine cutoff and their morning-light slot. They can also say that hygiene is not the whole program.

Module E. Diary how-to

Week: Starts Week 0. Continues every morning through Week 6 and any follow-up. Lineage: Carney et al. 2012, SLEEP, "The Consensus Sleep Diary: Standardizing Prospective Sleep Self-Monitoring." Official site: https://consensussleepdiary.com/ . Author page with for-profit permission note: https://drcolleencarney.com/sleep-diary/

Learning outcome. The client fills the diary every morning in under five minutes, without overnight logging, and treats the diary as the record that beats a wearable.

Field set (conceptual, not official item wording). Until Steuben clears for-profit use of the official CSD, teach these domains. Do not paste the official CSD form into a paid product.

• Time got into bed
• Time started trying to sleep
• Estimated minutes to fall asleep (SOL)
• Number of awakenings (not counting the final rise)
• Estimated minutes awake after sleep onset (WASO)
• Time of final awakening
• Time out of bed for the day
• Sleep quality rating (the CSD core includes a quality item)
• Optional comment: alcohol, caffeine late, nap, out-of-bed drill, illness, travel

Derived weekly (coach computes): TIB, TST, SE, mean SOL, mean WASO, mean awakenings, mean quality.

How to fill.

• Fill in the morning, once, after the final rise.
• Guess in round numbers. Precision theater is a form of clock-watching.
• Do not keep a notepad running at 03:00. Overnight logging wakes people and trains clock-watching.
• If they wear a device, they may glance at it for curiosity. They still write the morning estimate. The coach uses the diary. If diary and wearable disagree, the diary wins for protocol decisions.

License note. Permission for not-for-profit use of the official CSD is described by the authors as granted if the diary is unaltered and Carney et al. 2012 is cited. Permission for industry or for-profit use is obtained by contacting the first author (Carney). This business is client-paid. Steuben must clear CSD use before the official form is shown to a paying client. Until then, a field-domain diary that we wrote ourselves is the safe draft. Default is NEVER-BUY on official CSD. Spoken paraphrases and our own field list only.

Client actions. 14 baseline mornings if possible, then daily. Screenshot or export weekly.

Coach actions. Score completeness first (mornings filled / mornings expected). A week with fewer than 5 fills cannot support a window change. Teach one missing field at a time.

Abort rules. Shared list. A client who will not diary cannot run Module C. They can still receive A, B, and D.

Success (process). 5 or more complete mornings per week. Fields sufficient to compute SOL, WASO, TST, TIB, SE. No overnight logging. Wearable not used as the record.

4. Week-by-week min-path skeleton (6 weeks)

Monash core mapping: window (C), stimulus (B), circadian rise time (held from Week 1), diary (E). Hygiene (the rest of Module D) is one week (Week 2). Hygiene is never the program.

After Week 6, optionals may start only when the core is stable (diary complete, rise time held, window rules understood, no hot screen). Optionals are not a prize for paying more in week 1.

5. OPTIONAL MODULES (listed only; not in the first-client path)

Each optional is one paragraph on when it may start. None of these ships in the first 5 to 10 clients.

O1. Cognitive / worry-time. May start only after Weeks 2 to 3 if the diary is complete, the window (or no-window track) is stable, and the client's main remaining complaint is pre-sleep cognitive arousal (planning, replaying, threat thoughts) despite following B. A scheduled 15 to 20 minute worry / planning slot is placed at least 2 hours before the earliest bedtime. This is education in putting the day down. It is not psychotherapy and not a diagnosis of an anxiety disorder. If content is trauma-heavy or the client wants treatment for anxiety or depression, refer.

O2. Scheduled wind-down. May start only when the client is keeping the window and still arriving at bedtime in work clothes, with screens on, with no buffer. A 30 to 45 minute dim, low-demand buffer is scheduled to end at the earliest bedtime. This is a timing scaffold. It is not a spa product and not the program.

O3. Relaxation / breathing as skill, not treatment. May start only when the client leaves the bed (Module B) and needs a portable down-shift for the chair, or when they ask for a body skill after the core is in place. Teach one slow-breathing or progressive-muscle sequence as a skill for the out-of-bed place. Do not sell it as treatment for insomnia, as a cure, or as a replacement for A to C. AASM 2021 gives relaxation a conditional recommendation as a clinician single-component therapy. We still keep it optional and secondary.

O4. Alcohol education. May start only when the diary comments show evening alcohol on multiple nights, or the client asks. Teach that alcohol can shorten the time to fall asleep and then fragment the second half of the night, and can worsen breathing pauses in people at risk (Roehrs and Roth 2001, Sleep Medicine Reviews). We do not run an alcohol-use program. Hazardous use is a refer.

O5. Napping rules. May start only for clients whose job or caregiving makes a zero-nap rule collapse the week, or whose sleepiness on a cleared window becomes unsafe. A planned early, short nap with a hard end time may be safer than an accidental late nap. Unplanned long naps remain a rule break on the min-path.

O6. Shift-work timing variant. May start only after a dedicated protocol is written and Steuben and Napoleon sign. Shift work is a circadian and safety problem. It is not a bedtime tweak. Not in min-path. Residual: high. Do not improvise this for client 3.

O7. Travel. May start only for a planned trip after the core is stable. Teach: protect rise time in the new zone as soon as feasible, morning outdoor light in the destination, keep stimulus-control rules in the hotel bed, do not "catch up" with a 12-hour TIB the first night back. No jet-lag cure claim.

O8. Partner / bed-sharing constraints. May start only when the partner's schedule, a shared alarm, or sex-timing conflicts break the window or the out-of-bed drill. Personalize place and rise-time logistics. Do not delete stimulus control. Do not force a partner to become a patient.

6. What is NOT in the program

• Supplements, herbs, melatonin protocols sold as our method, or "stack" advice. Clients may talk to their own licensed clinician. We do not manage that.
• Wearables as diagnosis, as apnea detection, or as the number that sets the window.
• Hygiene-only packages, bedroom makeovers as the product, or 20-tip lists sold as CBT-I-informed coaching.
• Outcome-of-asleep guarantees: 8-hour nights, falling asleep in 15 minutes, cure, remission rates we did not measure.
• In-person labs, required polysomnography as our service, or medical-device positioning.
• Medication changes. We record names. We do not start, stop, or taper drugs.
• A second core path for the first 5 to 10 clients.

7. Coach weekly note (template)

Use this structure. Keep it process-first.

1. Diary completeness (n mornings / 7). Missing fields.
2. Rise-time spread (earliest to latest).
3. Stimulus-control: out-of-bed used? clock-watching? bed contents?
4. Window (if any): TIB clock hours, mean TST, SE, SOL, WASO.
5. Direction versus last week (up / down / flat). No promise language.
6. Safety: sleepiness, driving, mood-elevation, breathing, pregnancy. Hot? Refer.
7. Decision: hold / expand +15 / abort window / refer / stay on A+B+D+E.
8. One instruction for next week. One only.

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.

Personalization tree

Lifestyle in. Module pack + window timing out. Same protocol.

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Date: 2026-08-20 Audience: Internal coaches. English only.

Personalization changes clock hours and constraints. Personalization does not invent a second protocol for the first 5 to 10 clients. We do not diagnose, treat, or cure. We do not call our work CBT-I therapy or treatment of insomnia disorder. The core (Modules A through E) is always present unless a red-flag branch stops the window or stops enrollment.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen. Stimulus control means rules that rebuild the bed as a cue for sleep. A somnologist is a licensed sleep-medicine doctor. TST is total sleep time. TIB is time in bed. SE is sleep efficiency (TST divided by TIB). WASO is minutes awake after sleep onset. ISI is the Insomnia Severity Index, a copyrighted questionnaire. STOP-Bang is a copyrighted apnea screening questionnaire. IRLS is a copyrighted restless-legs scale. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. RoPA is the record of processing activities, the legal inventory of what personal data we store. PAP is a prescribed breathing device. RLS means a restless-legs pattern.

1. Fixed rules (read before the tree)

1. Personalization never deletes stimulus control. The five Bootzin-lineage rules stay. The place used for the out-of-bed drill may change (chair, hallway, another room). The rule does not disappear.
2. Personalization never deletes the diary. A client who will not fill a morning diary cannot run a window. They can still receive education and circadian modules if Steuben allows.
3. Personalization never substitutes hygiene-only. Module D may be emphasized when caffeine or light is an obvious breaker. Module D does not become the program.
4. Personalization never starts a window without a cleared screen. A beautiful rise-time plan does not override a hot item.
5. Red-flag branch is success. Any hot screen: refer, stop the window track, do not enroll in the sleep-window track. Coaching on A, D, and E only continues if Steuben allows. Otherwise full pause.
6. Output is a module pack plus clock hours. The pack for the first cohort is always A+B+C+D+E if the screen is clear, or A+B+D+E if the window is contraindicated and Steuben allows a limited track. Optionals stay closed until the min-path is stable.
7. Fit-call first gate is the spoken screen (03 section 17). Bedell writes the funnel. Monash writes the offer. They read the screen aloud. Do not invent a stored health quiz. Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD. Spoken paraphrases only. Until counsel residual + consent + RoPA row, the only loggable fields are (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no. No item-level notes.
8. Timezone default is Asia/Seoul (KST, UTC+9). If the client lives in another zone, convert every clock hour to that zone and write the zone on the window card. Do not keep two unofficial clocks.

2. Inputs (collect in Week 0, after or with the safety screen)

Collect these as logistics. They are not diagnoses.

Do not collect extra medical history here beyond what 03-intake-screen-content.md already covers. Duplicate questions create diagnostic theater.

3. Decision tree in prose

Step 0. Safety screen first. On the fit call, Bedell / Monash read the spoken screen in 03-intake-screen-content.md section 17. They do not invent a stored health quiz. They do not write down or store answers until Steuben issues certificates. The only mark is FIT or REFER. After certificates, the Steuben-wrapped form (sections 1 to 16) is the stored screen.

• If any item is hot: go to RED-FLAG BRANCH. Do not continue to window timing.
• If the screen is clear: continue.
• If the screen is unclear (client skipped items, or answers conflict): do not set a window. Ask once. If still unclear, treat as hot and refer.

Step 1. Diary gate. Is a 7-morning diary possible this week?

• No: enroll only in A + E (education and diary how-to). Reassess in 7 days. No window.
• Yes: continue. The first-cohort pack is A+B+C+D+E if later steps allow C.

Step 2. Window eligibility (cleared screen only). Ask whether any abort condition exists even if the form was "clear": pregnancy / trying / postpartum, mania history newly mentioned, seizure history, safety-critical work, already-unsafe sleepiness.

• Yes: no-window track. Pack = A+B+D+E. Referral still sent if the condition is medical.
• No: C is allowed after a usable baseline.

Step 3. Fix the rise time (the only timing decision that comes first). Compute the earliest rise time the client can keep 7 days a week.

• Work-bound: rise time = work start minus commute minus 45 to 60 minutes of dress / food / morning light.
• Caregiving-bound: rise time = the time they must be up for the other person, if that time is stable. If night caregiving is chaotic (multiple unpredictable wakes), do not set a tight window. Stay on A+B+D+E until wakes have a pattern or a licensed professional has cleared a plan.
• Free schedule (rare in a paid working client): rise time = a time they will actually keep, biased toward their stated preference, still fixed 7 days.
• Partner conflict: the paying client's rise time still must be fixed. A partner who sleeps later does not authorize a late weekend rise. Out-of-bed place is chosen so the partner is disturbed as little as possible.

Write the rise time as HH:MM ZONE. Example: 06:30 KST.

Step 4. Compute TIB from diary TST, then place bedtime. TIB = mean diary TST, with the floor and abort rules in Module C.

• If mean TST is at or above about 5.5 to 6.0 hours: TIB ≈ mean TST (round to 15 minutes).
• If mean TST is below about 5.5 hours: do not force a 5-hour window. Abort restriction. Refer or stay on A+B+D+E.
• Bedtime = rise time minus TIB.
• If that bedtime collides with work (client still on a late shift at the computed bedtime): do not steal sleep from a duty period. Either the rise time must move later (if mornings allow) or the window is not runnable this week. Shift-work is Optional O6 and is not in the min-path. For the first cohort, a client whose duty hours make a stable nocturnal window impossible is a no-window track plus referral if sleepiness is high.

Step 5. Apply lifestyle overlays (timing and constraints only).

Step 6. Module pack.

Step 7. Re-entry after referral. A client who returns with written clearance from a licensed professional is re-screened. The old window is not automatically restored. Restart from Step 0.

4. Red-flag branch (success, not a lost sale)

Trigger: any hot item on the screen, or a hot item that appears later.

Coach actions (process, not a diagnosis):

1. Stop any plan to set or continue a restricted window.
2. Do not tell the client they have apnea, RLS, bipolar disorder, or any disease.
3. Use the client-facing lines in 03-intake-screen-content.md. Core idea: this program is not a fit for the sleep-window track until a licensed professional has cleared these questions.
4. Refer to a somnologist or other licensed professional. Keep a short internal list of English-capable referral types (sleep medicine physician, psychiatrist, neurologist, OB) without naming a specific clinic in this draft. Slim runs delivery. Steuben is legal. They own the public referral list.
5. Log the event as a successful screen. Davout owns effort tracking and may still count the effort of a clean refer. Jomini counts a clean refer as a science win.
6. Do not offer a discount to "just do hygiene" as a consolation prize that replaces the refer.

5. Mermaid flowchart

flowchart TD
    start[Week 0 intake] --> screen{Safety screen}
    screen -->|Any hot item| red[RED FLAG: refer]
    red --> stopwin[Stop window track]
    stopwin --> steuben{Steuben allows limited coaching?}
    steuben -->|Yes| limited[Pack: A + D + E only]
    steuben -->|No| pause[Full pause. Referral is success]
    screen -->|Unclear after one ask| red
    screen -->|Clear| diary{7-morning diary possible?}
    diary -->|No| edu[Pack: A + E. Reassess in 7 days]
    diary -->|Yes| winok{Window safe? No pregnancy, mania, seizure, safety-critical work, high sleepiness}
    winok -->|No| nowin[Pack: A + B + D + E. No window]
    winok -->|Yes| rise[Fix rise time from work, commute, caregiving, zone]
    rise --> tst[TIB from diary TST]
    tst -->|TST below about 5.5h or client unsafe| abortc[Abort restriction. Refer or nowin]
    tst -->|TST ok| place[Bedtime = rise minus TIB]
    place --> collide{Duty hours collide with bedtime?}
    collide -->|Yes and shift-like| nowin
    collide -->|No| overlays[Apply caffeine, exercise, partner, light overlays]
    overlays --> core[Pack: A + B + C + D + E]
    core --> hold[Weeks 2 to 3: hold window]
    hold --> se{Weekly SE and safety}
    se -->|Unsafe sleepiness or new hot item| red
    se -->|SE high, safe, rise held| expand[+15 min TIB]
    se -->|SE mid or TST flat weeks 2 to 3| hold
    expand --> week6[Week 6 consolidate]
    hold --> week6
    nowin --> week6
    week6 --> opt{Min-path stable?}
    opt -->|Yes| optional[One listed optional may start only when the min-path is stable]
    opt -->|No| stay[Stay on core. No optional]

6. Worked timing examples (fictional, no real clients)

Example 1. Seoul office, commute, cleared screen. Work 09:00 KST, commute 50 minutes, wants 20 minutes of outdoor light. Rise time 07:00 is too late. Rise time 06:20 KST. Mean diary TST 6.0 h. TIB 6.0 h. Earliest bedtime 00:20 KST. Caffeine cutoff 16:20 KST (8 hours) or no later than 18:20 KST (6 hours). Pack: A+B+C+D+E.

Example 2. Partner sleeps later, cleared screen. Same numbers as Example 1. Out-of-bed place is the living-room chair. Sex is allowed in bed. Afterward the rules resume. Weekend rise stays 06:20 KST. Pack unchanged.

Example 3. Hot STOP-Bang-domain cluster. Loud snoring, witnessed pauses, high sleepiness. Red-flag. No window. Refer. Do not say "you have apnea." Pack: pause, or A+D+E if Steuben allows.

Example 4. Safety-critical work, otherwise "clear" form. Professional driver. Red-flag for the window. Refer for sleepiness and safety. Pack: no window. Do not run restriction that increases somnolence (Kyle et al. 2014).

Example 5. Mean TST 4.8 hours, screen clear. Do not set TIB to 5.0 and "see how it goes." Abort restriction. Refer or stay on A+B+D+E. Short diary TST can mean several things. We do not sort those causes.

Example 6. Night-shift nurse. Min-path window is built for a stable nocturnal night. This is Optional O6. First cohort: no-window track, education on light and caffeine only as Module D, refer if sleepiness is high. Do not improvise a 12-hour flip.

7. What the coach writes on the window card

Zone:
Rise time (7 days):
Earliest bedtime:
TIB (hours):
Mean diary TST used:
Screen: CLEAR / HOT (if HOT, stop)
Abort card given: yes / no
Caffeine cutoff:
Morning light slot:
Out-of-bed place:
Pack: A B C D E  (circle those in force)
Optionals: closed until the min-path is stable

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.

Intake screen: science content

Safety screen for whether behavioral coaching is appropriate

Status: SCIENCE CONTENT DRAFT. DRAFT-INTERNAL. Legal (Steuben) signs the form. Jomini writes the domains, the hot-screen meaning, and the educational copy. Date: 2026-08-20 Owner: Jomini writes sleep science. This is not a medical device. This screen produces no diagnostic output to the client.

Do not collect real client data against this draft. Do not publish. Do not paste copyrighted instrument items into any client-facing page until Steuben holds a written license.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen. A somnologist is a licensed sleep-medicine doctor. STOP-Bang is a copyrighted apnea screening questionnaire. ISI is the Insomnia Severity Index, a copyrighted questionnaire. IRLS is a copyrighted restless-legs scale. IRLSSG is the group that published restless-legs criteria. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. RoPA is the record of processing activities, the legal inventory of what personal data we store. RLS means a restless-legs pattern. PAP is a prescribed breathing device. TIB is time in bed. OSA means obstructive sleep apnea, a medical diagnosis we do not make.

1. Preamble (client-facing, educational)

Use language at this strength. Steuben may wrap, date, and jurisdiction-stamp it. Steuben may not turn it into a diagnosis engine.

This form is a safety screen. It helps us decide whether a research-backed sleep education and habit protocol, including a possible sleep-window, is appropriate for you. It is not a medical examination. It does not diagnose any sleep disorder or any other condition. We do not treat disease and we do not claim to cure. Some questions come from domains used in well-known screening instruments. Those instruments are copyrighted. Official item wording will appear only after our legal counsel has a license. Until then you will see plain-language questions about the same topics. If any answer suggests that a licensed professional should see you first, we will pause the sleep-window track and ask you to get clearance. That is a successful use of this form. It is how we keep the protocol inside its scope. You can skip a question. A skipped safety question is treated as uncleared. We will not start a sleep-window until it is answered or you have spoken with a licensed professional.

Coach-facing note on the preamble: the screen decides track, not disease. Hot means refer. Clear means the window may be considered after a diary baseline. Clear never means "you are healthy."

2. Instrument license gate (read before writing a form field)

Verbatim items: Legal must obtain license before client-facing use. That sentence stays on every domain block below.

Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD. Spoken paraphrases only.

3. STOP-Bang DOMAIN coverage

Purpose of this block: catch a cluster that should be seen by a somnologist before anyone shortens time in bed.

Source (name it on the form): STOP-Bang questionnaire, Chung and colleagues. Official site: http://stopbang.ca/ . Primary paper: Chung F, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812-821. https://doi.org/10.1097/ALN.0b013e31816d83e4 . Later scoring papers include Chung F, et al. High STOP-Bang score indicates a high probability of OSA. Br J Anaesth. 2012;108:768-775.

Copyright: Property of University Health Network. Academic licenses exist. Commercial use needs UHN / stopbang.ca permission. See also http://stopbang.ca/osa/screening.php for the official wording we must not copy.

Verbatim items: Legal must obtain license before client-facing use.

3.1 Domains (educational paraphrase only)

Ask in our own words. Yes / no / skip. Do not number them as an official STOP-Bang score on the client page.

1. Snoring. Habitual, loud snoring, the kind a partner comments on or that is obvious through a door. We are asking about a pattern, not a single cold.
2. Tired / sleepy. Frequent daytime tiredness, fatigue, or sleepiness, including a tendency to doze when sitting or, more seriously, when driving or talking.
3. Observed breathing pause. Someone has seen the client stop breathing, choke, or gasp during sleep.
4. Pressure. Known high blood pressure, or current treatment for high blood pressure.
5. Body size. A body-mass index in a range the official instrument treats as a positive bang item (the official threshold is published on stopbang.ca). Until licensed, ask height and weight as optional logistics, or ask whether a clinician has already flagged a high BMI. Do not present a home BMI calculator as a diagnostic device.
6. Age. Age in a range the official instrument treats as a positive bang item (official site: older than 50).
7. Neck. A large neck / shirt-collar size in the range the official instrument uses (official site publishes a collar threshold). Ask "has a clinician or tailor remarked on a large neck size?" rather than claiming we measured it.
8. Gender. The official instrument treats male gender as a bang item. Ask gender as the client describes it. Do not lecture. Do not imply that women are "safe."

3.2 Coach-facing scoring intent (never shown to the client as a disease label)

• Official STOP-Bang scoring (0 to 8, with published low / intermediate / high bands on stopbang.ca) is licensed territory. Do not print those bands on a client PDF until Steuben has the license and a clinician-overseen interpretation policy. We are not that clinic.
• Hot (process meaning): two or more of the first four domains (snoring, sleepy, observed pause, pressure) or an observed pause / gasping or sleepy-plus-snoring or any already-diagnosed apnea / current PAP. Also treat a dense cluster across bang-type domains (body size + age + neck + male + snoring) as hot even if we are not emitting an official score.
• Hot means: do not start a sleep-window. Refer to a somnologist or licensed sleep professional. Write the client line in section 12. Do not say "you have sleep apnea" or "you scored high-risk OSA."
• Clear-enough for window consideration: no observed pauses, no PAP, no dense snoring-plus-sleepiness cluster, remaining items sparse. Still not a medical all-clear.

Chung 2008 reported that adding BMI, age, neck, and gender to STOP raised sensitivity against polysomnographic AHI cutoffs in a preoperative sample. That is why we cover eight domains. It is also why we do not pretend a home form replaces a sleep study.

4. IRLSSG question-set DOMAINS

Purpose of this block: catch an RLS-pattern that should be medically reviewed before we restrict time in bed or blame "restlessness" on conditioned arousal.

Source: Allen RP, et al. Restless legs syndrome / Willis-Ekbom disease diagnostic criteria: updated IRLSSG consensus criteria. Sleep Medicine. 2014;15(8):860-873. https://doi.org/10.1016/j.sleep.2014.03.025 . Europe PMC abstract live-verified: https://europepmc.org/article/MED/25023924 . Group site commonly cited: https://www.irlssg.org/

License note: The IRLSSG criteria are a published consensus description. The IRLS severity rating scale and related instruments are copyrighted by the IRLSSG and, for many uses, licensed through Mapi / ePROVIDE (fees for commercial users). Do not paste IRLS items. Do not produce an IRLS severity class.

Verbatim items: Legal must obtain license before client-facing use of any official IRLS / sIRLS form.

4.1 Domains (educational paraphrase)

Ask in our own words. The 2014 update keeps the classic four features and adds a fifth: the pattern is not better explained solely by another condition.

1. Urge to move. A need to move the legs (sometimes the arms), often with uncomfortable sensations.
2. Rest. The urge begins or grows while the person is at rest (sitting, lying).
3. Relief. Movement (walking, stretching) eases the urge, at least while they are moving.
4. Evening. The pattern is worse in the evening or night than in the morning.
5. Not solely another cause. The pattern is not explained only by a cramp, a numb position, swelling-pain, arthritis pain, or ordinary fidgeting. Allen 2014 added this differential-diagnosis criterion to reduce false labels.

4.2 Coach-facing scoring intent

• Hot (process meaning): the client endorses an evening urge-to-move that starts at rest and eases with movement, or they already carry an RLS diagnosis. Also hot if they describe painful restlessness that keeps them walking the room most nights and they want us to "fix it with a shorter window."
• Hot means: refer to a somnologist, neurologist, or other licensed professional. Do not start a sleep-window. Do not tell them they have RLS. Do not start iron, gabapentin, or any other agent.
• A single "my legs ache after sport" is not hot. A positional numb foot that resolves by turning over is not hot. When unsure, treat as uncleared and refer.

5. Breathing / witnessed apneas / gasping (beyond STOP-Bang)

Ask explicitly, in our words:

• Has anyone seen you stop breathing, choke, or gasp during sleep?
• Do you wake with a choke, a snort, or a sense of air-hunger?
• Do you wake with a dry mouth, morning headache, or a partner who leaves the room because of breathing noise? (supporting context, not a score)
• Do you use PAP, oxygen, a dental sleep appliance, or positional therapy prescribed by a clinician?

Hot: any witnessed apnea, recurrent gasping, or current PAP / oxygen / prescribed appliance. Action: refer. If they are already under a sleep physician and stable on PAP, still do not start a restricted window unless that physician has cleared it in writing. Untreated or unstable breathing problems plus a shorter TIB is a safety error.

6. High sleepiness / drowsy driving / safety-critical work

This block is ours. It is not an ESS. If we ever want an official ESS score, Steuben licenses Johns / Mapi first.

Domains to ask:

• Sleepiness that makes it hard to stay awake while sitting (desk, meeting, screen).
• Any drowsy driving, lane drift, or near-miss in the last year.
• Sleepiness while caring for a child, cooking, or using tools.
• Job category: pilot, air-traffic, professional driver, surgeon / proceduralist, heavy machinery, night security, emergency response, and any role the client flags as safety-critical.

Hot: any drowsy driving or near-miss; sleepiness in a safety-critical role; nodding off in unsafe settings. Action: refer. No window. Kyle et al. 2014 (Sleep) showed sleep-restriction protocols can increase somnolence and impair vigilance. We do not add that load to an already sleepy or safety-critical client.

Coach-facing: "hot" here means unsafe to restrict, not "this person has hypersomnia" as a label.

7. Bipolar / mania / hypomania history

Ask in plain language:

• Has a licensed professional ever diagnosed bipolar disorder, mania, or hypomania?
• Has there been a stretch of needing much less sleep and feeling unusually energetic, wired, or unstoppable?
• Is a psychiatrist or other prescriber currently involved?

Hot: any yes, or a vivid self-description of reduced-need-for-sleep plus elevated mood. Action: no window. Refer to the existing prescriber or to a licensed psychiatrist. Sleep loss is a known risk in this population. Harvey-group CBT-I-BP work used a higher TIB floor and close mood monitoring in a clinical trial. We are not that trial and we have no clinician on staff.

If they are already in psychiatric care and want education only, Steuben decides whether A+D+E may run. Jomini still forbids a restricted window without written specialist clearance.

8. Seizure history

Ask:

• Any seizure, epilepsy diagnosis, or a clinician warning that sleep loss is a risk for you?

Hot: any yes. Action: no window. Refer to the treating neurologist or a licensed professional. Sleep deprivation can lower seizure threshold. We do not run that experiment.

9. Pregnancy / trying to conceive / postpartum (window abort)

Ask:

• Currently pregnant?
• Trying to conceive?
• Given birth in the last 12 months, or currently feeding a child at night?

Hot for the window: any yes. Action: do not start a restricted sleep-window. Refer to OB / primary / a licensed professional for sleep and medical questions. Night feeding makes a tight window dishonest anyway. Education on stimulus control (adapted), diary, and circadian timing may continue if Steuben allows.

10. Existing sleep-disorder diagnosis or current PAP / oxygen

Ask them to name, in their words, any sleep diagnosis they have already been given (apnea, narcolepsy, idiopathic hypersomnia, periodic limb movements, parasomnia, circadian disorder, and others) and any device they use.

Hot for the window: apnea, narcolepsy / hypersomnia, REM-behavior or other parasomnia with injury, any oxygen or PAP. Action: refer back to the treating clinician. We do not take over device management.

A prior "insomnia" label from a GP, without the above, is not automatically hot. It is also not a diagnosis we confirm.

11. Substances: sedative-hypnotics and stimulants

Ask for a name-and-timing list: prescribed sleep medication, benzodiazepines, Z-drugs, dual-orexin antagonists, sedating antidepressants used at night, stimulants, wake-promoters, and non-prescribed sedatives.

Rules:

• We do not manage medications. We do not start, stop, or taper.
• A client on a sedative-hypnotic can still run A, B, D, and E if the screen is otherwise clear. A window is a case-by-case Steuben / safety decision. Default: do not add restriction on top of a sedative without the prescriber's knowledge. Send a one-page "your patient is starting a behavioral sleep-education protocol" note if the client consents. Steuben drafts that note.
• Stimulants late in the day are a Module D timing issue. They are not ours to discontinue.
• Hazardous alcohol or other substance use is a refer (see Optional O4). We are not an addiction clinic.

12. Hard-stop rule

Any hot item:

1. Do not start a sleep-window.
2. Refer to a somnologist or other licensed professional.
3. Coaching may continue only on education / diary / circadian modules if Legal (Steuben) allows.
4. Otherwise, full pause.

Uncleared (skipped) safety items are treated as hot until answered.

A later-emerging hot item (new pregnancy, new gasping, new drowsy driving, new elevated mood with no sleep) aborts a window that was already running. Widen first if they are acutely sleepy, then refer.

13. Coach-facing: what "hot" means

Internal tags: HOT-BREATHING, HOT-RLS, HOT-SLEEPY, HOT-SAFETYJOB, HOT-MOOD, HOT-SEIZURE, HOT-PREG, HOT-EXISTING. These tags are for the coach file. They are not printed on the client receipt.

14. Client-facing output copy examples

Use these. Do not improve them into a diagnosis.

On any hot item:

"This program is not a fit until a licensed professional has cleared these questions. We are pausing the sleep-window part of the protocol. That is the safe use of this screen. It is not a diagnosis."

On breathing / observed pause cluster:

"Your answers include breathing or sleepiness questions that sit outside a behavioral coaching program. Please see a somnologist or other licensed clinician before we consider any sleep-window. We can share a short list of the topics you may want to mention. We will not label a disease from this form."

On RLS-pattern cluster:

"Your answers include an evening urge-to-move pattern that should be reviewed by a licensed professional before we change time in bed. This form does not diagnose restless legs syndrome."

On sleepiness / safety-critical work:

"Because of sleepiness or the safety demands of your work, we will not shorten your time in bed. Please have a licensed professional review this. We can continue, if counsel allows, with education, a diary, and clock-timing habits only."

On a clear screen (never a health certificate):

"Nothing on this screen blocks a possible sleep-window later. That is not a medical all-clear. If new symptoms appear (gasping, drowsy driving, pregnancy, a big mood change, a seizure), tell us and we will pause."

On a skipped item:

"A skipped safety question means we cannot start a sleep-window. You can complete the item or take the question to a licensed professional."

15. Form fields Steuben must wrap

Jomini does not write counsel's clauses. Steuben must add, at minimum:

1. Consent to a non-diagnostic educational screen and to CBT-I-informed behavioral sleep coaching (named that way).
2. Acknowledgement that we do not diagnose, treat, or cure, and that we are not a medical device, not a clinic, and not a substitute for a physician.
3. Data: what is stored, where, who sees it, retention, deletion. No real client data in this repository.
4. Jurisdiction: business operated from Seoul, Republic of Korea. Clients are English-speaking and online. Steuben states which law governs, how KR medical-advertising rules apply to a coaching offer, and how foreign clients are handled.
5. Instrument licenses: a checkbox that verbatim ISI / STOP-Bang / IRLS / ESS items will appear only if a license is on file. Until then, domain paraphrases only.
6. Referral consent: permission to tell the client to see a licensed professional, and optional permission to send a one-page note to a named clinician.
7. Medication clause: we do not manage drugs.
8. Emergency: drowsy driving, suicidal ideation, chest pain, and similar go to local emergency services. We are not an emergency service. (If suicidal ideation is asked, Steuben and a licensed consultant own that item. Jomini does not turn this form into a psych crisis tool.)
9. Signature, date, timezone, English-language confirmation.

16. What this form must never do

• Emit a disease name as our finding.
• Emit an official instrument total that we are not licensed to compute.
• Auto-enroll a hot client into a "lite" paid window.
• Store a wearable AHI-like score as if it were a diagnosis.
• Promise that a clear screen means the client will sleep well.

17. Fit-call spoken screen (Bedell / Monash read-aloud)

Napoleon routing 2026-08-20: Monash writes the offer and uses min-path modules (window, stimulus, circadian, diary; hygiene one week). The fit call must not invent a stored health quiz. Screen-to-refer only. Answers are not stored until Steuben issues certificates.

SPOKEN SCREEN RULES (boxed; read before the call)

• Bedell / Monash read these questions verbatim on the fit call.
• Do not invent extra health-quiz items.
• Do not write, type, record, screenshot, CRM-log, or otherwise STORE answers until Steuben issues certificates.
• Screen-to-refer only. A refer is a successful call.
• Do not score STOP-Bang. Do not name a disease. Do not compute BMI on the call.
• If ANY answer is HOT: stop asking further health questions, speak the refer line, do not sell the sleep-window track.
• Education-only / diary / circadian continuation after a refer is a Steuben decision, not a closer decision.
• Do not add unofficial STOP-Bang scoring items (BMI, neck, gender-as-risk) to this spoken set. Those stay behind the license stop rule and the storage stop rule.

One-line map: Spoken Qs cover hard-stop domains. They are NOT a licensed STOP-Bang, IRLS, ESS, or ISI.

Preamble (Read verbatim)

I am going to ask a short safety screen out loud. This is not a medical exam. We do not diagnose and we do not treat. I will not write down or store your answers until our legal certificates are in place. If any answer means the sleep-window part of this protocol is not a fit, I will tell you that and point you to a licensed professional. That is the correct use of this screen.

Questions (read verbatim, Yes / No)

Q1. Has a doctor or sleep clinic already told you that you have sleep apnea, or do you use a CPAP, oxygen, or a prescribed sleep-breathing device? HOT if yes. Action: refer somnologist / treating clinician. No window. Stop the remaining health questions.

Q2. Has anyone seen you stop breathing, choke, or gasp in your sleep? HOT if yes. Action: refer somnologist. No window. Stop the remaining health questions.

Q3. Do you snore loudly on most nights, and also feel very sleepy in the daytime? HOT if yes to the combination. Action: refer somnologist. No window. Stop the remaining health questions.

Q4. In the last year, have you had drowsy driving, a near-miss, or nodding off in an unsafe situation? HOT if yes. Action: refer licensed professional. No window. Stop the remaining health questions.

Q5. Is your work safety-critical if you were sleepy? Examples: pilot, professional driver, surgeon, heavy machinery, night security, emergency response. HOT if yes. Action: refer. No window. Stop the remaining health questions.

Q6. Do you get an urge to move your legs in the evening that starts when you rest and eases when you walk or stretch? HOT if yes. Action: refer somnologist or neurologist. No window. Do not say "you have restless legs syndrome." Stop the remaining health questions.

Q7. Has a licensed professional ever diagnosed bipolar disorder, mania, or hypomania? Or have you had a stretch of needing much less sleep and feeling unusually wired or unstoppable? HOT if yes to either part. Action: refer psychiatry / existing prescriber. No window. Stop the remaining health questions.

Q8. Have you ever had a seizure, or has a clinician warned you that sleep loss is a risk for you? HOT if yes. Action: refer neurology / treating clinician. No window. Stop the remaining health questions.

Q9. Are you pregnant, trying to conceive, or in the first year after giving birth? HOT if yes. Action: no window. Refer OB or licensed professional. Stop the remaining health questions.

Q10. Has a clinician given you any other sleep-disorder diagnosis besides ordinary insomnia, such as narcolepsy or a parasomnia with injury? HOT if yes. Action: refer treating clinician. No window. Stop the remaining health questions.

If the person volunteers extra medical detail, thank them. Do not write it down. If it sounds like a hot domain, treat as REFER.

Refer line (read verbatim on any HOT)

This program is not a fit until a licensed professional has cleared these questions. We will not start a sleep-window. That is the safe use of this screen. It is not a diagnosis.

Clear line (only if every answer was No)

Nothing you said on this call blocks a possible sleep-window later. That is not a medical all-clear. If anything on this list changes, tell us and we pause.

No-store SOP (Steuben-encoded, 2026-08-20)

STOP RULE (Steuben via Napoleon, 2026-08-20).

Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD.

Spoken paraphrases only.

No-store SOP until counsel residual + consent + RoPA row: the only loggable fields are (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no.

No item-level notes.

Do not invent KR advertising copy.

SCIENCE CONTENT DRAFT. DRAFT-INTERNAL. Steuben signs the form. Legal has not signed. No public assets. No client data.

Happy-client definition: sleep-number half

Official ISI unused (never-buy). Field-domain morning log. Direction from literature. No asleep guarantees.

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Date: 2026-08-20 Scope: Sleep-number half only. Davout owns effort tracking. This file does not write that half. No-asleep-guarantee rule: Monash writes the offer. Bedell writes the funnel. They may not turn any number in this file into an outcome-of-asleep guarantee.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. ISI is the Insomnia Severity Index, a copyrighted questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. SOL is minutes to fall asleep. WASO is minutes awake after sleep onset. NWAK is the number of awakenings. TST is total sleep time. TIB is time in bed. SE is sleep efficiency (TST divided by TIB). A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen.

1. What this half is for

A "happy client" in this business has two halves. Davout scores whether the client ran the protocol (diary, rise time, out-of-bed rule, window held). Jomini scores whether the sleep numbers we are allowed to watch moved in the direction published reviews report, without promising a destination.

This half exists so a coach does not:

• declare failure because TST is flat in weeks 2 to 3,
• declare victory because a wearable said "8 hours,"
• or hand Monash a fake remission rate.

2. Instruments

2.1 Insomnia Severity Index (ISI)

• What it is: A brief self-report of perceived insomnia severity and daytime impact. Developed by Charles M. Morin. Validation: Bastien, Vallières, and Morin 2001, Sleep Medicine, 2(4):297-307. https://doi.org/10.1016/S1389-9457(00)00065-4
• Copyright: © Morin, C.M. (1993, 1996, 2000, 2006). Distributed by Mapi Research Trust / ePROVIDE. Commercial and many funded uses need a license. Live-verified: https://eprovide.mapi-trust.org/instruments/insomnia-severity-index (fetched 2026-08-20).
• License residual: CLOSED as NEVER-BUY (Steuben via Napoleon, 2026-08-20). Official ISI unused. See RESIDUALS.md.
• Our use: NEVER-BUY. Official ISI unused. Do not administer ISI. Never-buy stop rule. Do not show ISI items. Do not collect an ISI total.
• Process pulse: A single plain-language "how much of a problem has sleep been this week, 0 to 4?" is allowed only if it is not named ISI. It is a process pulse. It is not an ISI.
• Direction in the literature: ISI scores tend to fall in CBT-I trials. That is a published direction. It is not our promise. Do not invent a point-drop target as a guarantee ("you will drop 8 points").
• Published numbers: Bastien 2001 showed the ISI is sensitive to change in a behavioral / pharmacologic trial in older adults. A later Morin et al. 2011 paper in SLEEP (34(5):601-608) discusses psychometric indicators and treatment response. I did not fetch a full-text page that states a single canonical mean ISI change for Trauer 2015 (Trauer reported diary SOL / WASO / TST / SE, not ISI, in the abstract I fetched). Numeric ISI effect size is not restated here pending a licensed copy and a full-text confirm of any specific mean change we might want to quote. Direction only: scores tend to fall.

2.2 Consensus Sleep Diary (CSD)

• What it is: Expert-consensus morning diary. Carney CE, Buysse DJ, Ancoli-Israel S, Edinger JD, Krystal AD, Lichstein KL, Morin CM. The consensus sleep diary: standardizing prospective sleep self-monitoring. SLEEP. 2012;35(2):287-302. https://doi.org/10.5665/sleep.1642
• Sites: https://consensussleepdiary.com/ (live-verified). https://drcolleencarney.com/sleep-diary/ (live-verified). Both state the diary is free for not-for-profit use if unaltered and cited, and that industry / for-profit use requires contacting the first author.
• License residual: CLOSED as NEVER-BUY (Steuben via Napoleon, 2026-08-20). Never-buy the official CSD form. Use the field-domain morning log we wrote (see Module E).
• How we collect: Morning only. No overnight logging. English. Async.

3. Metrics we watch (from the diary)

Compute weekly means from mornings actually filled. Do not impute from a wearable.

A week with fewer than 5 fills: report completeness only. Do not make a hold / expand decision on 3 noisy nights.

4. Expected DIRECTION from literature (not promises)

Primary source fetched as an abstract (Europe PMC, live 2026-08-20): Trauer JM, Qian MY, Doyle JS, Rajaratnam SMW, Cunnington D. Cognitive Behavioral Therapy for Chronic Insomnia: A Systematic Review and Meta-analysis. Annals of Internal Medicine. 2015;163(3):191-204. https://doi.org/10.7326/M14-2841 . PMID 26054060. Abstract URL: https://europepmc.org/article/MED/26054060

Trauer included 20 face-to-face multimodal CBT-i trials (1,162 participants) versus inactive comparators, in chronic insomnia without the comorbidities they excluded. At the post-treatment time point the abstract reports:

Trauer also states changes seemed sustained at later time points and that no adverse outcomes were reported in the included trials. Their limitation: narrow inclusion, less clear later-time estimates.

AASM 2021 (Edinger et al., J Clin Sleep Med, 17(2):255-262, https://doi.org/10.5664/jcsm.8986 ) and ACP 2016 (Qaseem et al., Ann Intern Med, https://doi.org/10.7326/M15-2175 ) recommend CBT-I as first-line for clinicians treating chronic insomnia disorder. We translate methods into coaching. We do not claim we reproduced those guideline outcomes. We do not claim we treat the disorder.

Client-facing allowed sentence (needs-legal if used in ads):

"In published reviews of CBT-I-style programs, people often take less time to fall asleep, spend less time awake in the night, and spend a larger share of the night actually asleep. Total sleep time often changes little at first. Those are research patterns. They are not a guarantee of what you will see."

Client-facing forbidden:

• "You will sleep 8 hours."
• "You will fall asleep in 15 minutes."
• "We cure insomnia."
• "X% of our clients remit." (We have no measured cohort yet.)
• "You will match Trauer's 19 minutes."

5. ISI direction (again, no point-drop guarantee)

• Official ISI unused. NEVER-BUY. Do not administer ISI. Never-buy stop rule. Do not use ISI as a repeated self-report.
• Direction: scores tend to fall in CBT-I trials (Bastien 2001; later Morin psychometric papers).
• Do not set a happy-client rule like "ISI must fall by 7 points." That would be an invented guarantee.
• If Steuben later wants a published mean change quoted, fetch the full text of the specific trial or review and add a ledger row. Until then: direction only.

6. What we do NOT claim

• 8-hour nights as a goal or a promise. Need varies. The protocol does not target a universal 8.
• Falling asleep in a set number of minutes.
• Cure, remission, or "insomnia disorder resolved."
• Remission rates, responder rates, or "X% of clients" from this practice. We have not measured them.
• Wearable "sleep score" improvement as a happy-client metric.
• Supplement response.
• That a clear intake screen predicts a good diary week.

Monash / Bedell: this list is the no-asleep-guarantee wall. If a landing page needs a number, it may use a Trauer sentence marked as published-review, not-our-promise, after Steuben reads it. It may not use a future-tense "you will."

7. Coach rule: weeks 2 to 3

Flat TST + rising SE = on-protocol. Hold the window.

This is the rule that protects the science from the client's (and the coach's) panic.

Why: restriction / a tighter window concentrates the sleep the client is already getting. SE rises because TIB shrank and WASO / SOL often shrink. TST often stays similar until later expands (Trauer's TST CI includes zero at post-treatment). Widening in week 2 to "give them more sleep" puts wake back in the bed and undoes B and C.

Write this in the Week 2 note before the client asks:

"If your total sleep time looks similar to last month and a larger share of the night is sleep, that is the pattern published reviews describe in the early stretch. We hold the window. We do not chase a longer night yet."

Expand only on the Module C table (about 90% SE, rise time held, sleepiness safe), and only in 15-minute steps.

8. How numbers are collected

Collection is async. English. The client fills. The coach computes. No overnight logging.

9. Happy-client sleep-number rubric (internal)

Use this in Week 6 reviews. It is a rubric for us. It is not a certificate for the client.

On-protocol (science half satisfied enough to call the number-side "in range"):

• Diary completeness ≥ 5 mornings in at least 4 of the 6 weeks, including Week 6.
• Rise-time spread usually inside about 60 minutes.
• SOL direction down or clearly less chaotic versus baseline, or WASO direction down, or SE direction up. One of those three is enough. We do not require all three.
• TST flat is acceptable. TST up is acceptable. TST down with unsafe sleepiness is a safety fail (abort), not a number fail.
• Process pulse 0 to 4 (not named ISI) may be noted as direction only. Do not administer ISI. Never-buy stop rule.

Off-protocol (do not call the number-side happy):

• Window widened in weeks 2 to 3 to chase TST.
• Wearable used as the record.
• Completeness too low to compute.
• Hot screen ignored.
• Any guarantee language used with the client.

Davout will add effort flags (did they leave the bed, did they keep the rise time). A client can be on-protocol for numbers and off-protocol for effort, or the reverse. Both halves are required before Napoleon calls the engagement a model case.

10. Sentences Monash / Bedell may request, and the answer

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data. Davout owns the other half.

Source ledger

Every client-facing sleep claim we may use

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Owner: Jomini writes sleep science. Date: 2026-08-20 Rule: A client-facing sleep claim with no row here does not ship. Monash writes the offer. Bedell writes the funnel. They draw from this list only.

Terms used in this file: CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy. A sleep-window is a planned stretch of time in bed, set from the diary after a clear screen. Stimulus control means rules that rebuild the bed as a cue for sleep. ISI is the Insomnia Severity Index, a copyrighted questionnaire. STOP-Bang is a copyrighted apnea screening questionnaire. IRLS is a copyrighted restless-legs scale. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. SOL is minutes to fall asleep. WASO is minutes awake after sleep onset. TST is total sleep time. SE is sleep efficiency. OSA means obstructive sleep apnea, a medical diagnosis we do not make.

URL marks:

• live = fetched or confirmed reachable on 2026-08-20
• unverified-live = canonical URL believed correct; this environment's fetch failed or returned a block / wrong page
• abstract-live = abstract page fetched; full text not confirmed

Strength: strong = guideline or consistent review support for the direction of the claim. conditional = real evidence, narrower or mixed. folklore-killed = common belief that evidence contradicts or does not support.

Client-facing allowed?: yes = educational wording in the Claim column may be used after Steuben's general copy pass. needs-legal = science is fine, advertising / KR scope / license must be checked. no = coach-facing only, or forbidden as written.

License residual (read first)

ISI, STOP-Bang, IRLS / IRLSSG instruments, and the Epworth Sleepiness Scale are copyrighted. Commercial reproduction of verbatim items likely needs a license (Mapi / ePROVIDE for ISI and ESS; University Health Network / stopbang.ca for STOP-Bang; IRLSSG / Mapi for IRLS). Consensus Sleep Diary is more usable and still needs first-author permission for for-profit use. This residual is HIGH. Do not paste official items. Domain paraphrases only until Steuben holds paper.

Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD. Spoken paraphrases only.

Ledger

L01. CBT-I-informed methods and SOL / WASO / SE

L02. TST often flat early

L03. AASM 2021 first-line behavioral recommendation (translated)

L04. ACP 2016 first-line CBT-I (translated)

L05. Sleep window / restriction (Spielman)

L06. Stimulus control (Bootzin)

L07. Hygiene alone is insufficient

L08. Circadian morning light (conditional)

L09. Caffeine timing (conditional)

L10. STOP-Bang as a screen, not a diagnosis

L11. IRLSSG domains as a screen, not a diagnosis

L12. ISI as a measure (not a promise)

L13. Consensus Sleep Diary as the record

L14. Restriction can increase somnolence (Kyle)

Folklore killed

F01. Warm milk cures insomnia

F02. Eight hours is mandatory

F03. Catch-up weekend sleep is equivalent

F04. Alcohol is a sleep aid

F05. Wearables diagnose apnea

F06. More time in bed always helps

F07. Trying harder produces sleep

F08. Hygiene list alone is the program

Claims we will not write (no row = no ship)

• Any future-tense asleep guarantee ("you will…").
• Any invented percentage of our clients.
• Any "we diagnose / treat / cure."
• Any supplement protocol as our method.
• Any "this wearable says you have apnea."
• Any official instrument total without a license.
• Any shift-work performance guarantee (O6 is not in min-path).

Fetch log (2026-08-20)

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.

Residuals and confidence

Sleep-science folder, 2026-08-20

Status: DRAFT-INTERNAL. Legal has not signed. Do not publish. Do not collect real client data. Owner: Jomini writes sleep science. Audience: Napoleon, Steuben, Jomini. Not a public document.

Terms used in this file: ISI is the Insomnia Severity Index, a copyrighted questionnaire. STOP-Bang is a copyrighted apnea screening questionnaire. IRLS is a copyrighted restless-legs scale. ESS is the Epworth Sleepiness Scale, a copyrighted sleepiness questionnaire. CSD is the Consensus Sleep Diary, a copyrighted morning diary form. RoPA is the record of processing activities, the legal inventory of what personal data we store. A somnologist is a licensed sleep-medicine doctor. CBT-I is the published clinical method called cognitive behavioral therapy for insomnia. We do not call our work CBT-I therapy.

1. Instrument-license risk. HIGH residual (first)

This is the residual most likely to create a legal defect if ignored.

License residual ranks (read first):

• ISI: NEVER-BUY (Steuben via Napoleon, 2026-08-20)
• STOP-Bang: NEVER-BUY
• IRLS: NEVER-BUY
• ESS: NEVER-BUY
• Official CSD: NEVER-BUY. Use the field-domain morning log we wrote.
• Spoken fit-call no-store SOP encoded 2026-08-20. Only loggable fields: (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no. No item-level notes. Further store waits on counsel residual + consent + RoPA row.
• Korean medical advertising / coaching-vs-medical: Steuben. Jomini does not invent KR advertising copy.
• No clinician on staff; medical-identity staffing is a failure mode
• Shift-work not in min-path
• Exact ISI mean change not restated (Trauer abstract is diary metrics)

Rule already in the job description: verbatim items: Legal must obtain license before client-facing use. This line is also in 03-intake-screen-content.md and 05-source-ledger.md.

Default is NEVER-BUY on official ISI, STOP-Bang, IRLS, ESS, and official CSD (Steuben via Napoleon, 2026-08-20). Keep spoken paraphrases. Never emit official scores. Never name a number as an ISI or STOP-Bang total. Reopen a buy path only if Napoleon and Steuben reverse.

2. Korean medical advertising / coaching-versus-medical scope. Hand to Steuben

• Founder in Seoul. Facing name: Emperor Livingstone. Legal name on contracts: 이성재. Clients English-online. Jurisdiction is mixed and is not Jomini's to invent.
• KR rules on medical advertising, use of the words for treatment / clinic / diagnosis, and cross-border coaching are Steuben's residual.
• This packet already forbids diagnostic output, cure claims, and "we deliver AASM therapy." That is necessary. It may not be sufficient under KR law.
• No clinician is on staff in this draft. Implying one is a failure mode (job description section 7).

3. Exact published effect sizes pending full-text

Rule honored: if a number was not on a page we fetched, we stated direction and wrote "numeric effect size not restated here pending full-text confirm" (ISI). Trauer diary numbers were on the fetched abstract, so they are quoted with CIs and a not-our-promise label.

4. Shift-work protocol not in min-path

Optional O6 is listed only. It may start only when a dedicated protocol is written and signed. A night-shift or rotating-shift client in the first 5 to 10 is a no-window track plus Module D caution, and a refer if sleepiness is high. Do not improvise a 12-hour flip for client 3. Writing a real shift protocol needs its own addendum, safety review, and Napoleon sign.

Travel (O7) and partner (O8) are listed the same way.

5. No clinician on staff

• Medical-identity staffing is a failure mode.
• Hot screens refer out. We do not keep a "house somnologist" in the copy unless one is actually hired and Steuben rebuilds scope.
• Creating a clinician helper, a diagnosis helper, or a supplements helper is forbidden (job description section 10).

6. Spoken fit-call no-store SOP (Steuben-encoded, 2026-08-20)

Napoleon ordered a spoken screen on the fit call. Section 17 of 03-intake-screen-content.md is the script. The no-store SOP is encoded. Until counsel residual + consent + RoPA row exist, the only loggable fields are (1) date (2) FIT or REFER (3) sale stopped yes/no (4) generic referral spoken yes/no. No item-level notes. Do not invent KR advertising copy.

7. Other residuals

• CSD arithmetic: TST working definition must stay consistent. Coaches need a one-page calculator (Slim). Not written as software in this folder.
• BMI / neck measurement: official STOP-Bang uses specific thresholds. We avoided building a diagnostic calculator. Height / weight as optional logistics still has medical-device and privacy taste. Steuben to trim.
• Suicide / crisis items: not designed here. If Steuben adds them, a licensed consultant owns the wording. We are not an emergency service.
• Medication letters to prescribers: Steuben drafts. Jomini will not write a taper.
• Wearable API temptation: Slim may not plug a ring into the window formula.
• First-cohort outcome math: no "X% of our clients" until a measured cohort exists and Steuben signs the statistic.
• Unverified-live URLs: listed in 05-source-ledger.md fetch log. Canonical citations still included.
• Language / style stop rule: these drafts were written to avoid em dashes, contrast slogans, and corporate verbs. A later editor must not put them back.

8. Confidence per deliverable (0 to 1)

Folder-level confidence: 0.87 that a coach can train internally from this packet tomorrow. 0.40 that any of it can go public, because Steuben has not signed licenses, the form, or KR scope.

9. What would raise confidence

1. CLOSED (2026-08-20). Instrument decision is NEVER-BUY for official ISI, STOP-Bang, IRLS, ESS, and official CSD. Reopen only if Napoleon and Steuben reverse.
2. Full-text confirm of Trauer later time points and of Drake 2013 if we want tighter caffeine copy.
3. A fictional end-to-end dry run (Week 0 to Week 6) with Napoleon and one coach, no real client.
4. KR advertising memo from Steuben. Jomini does not invent KR advertising copy.
5. Davout's effort-half document, so the happy-client definition is complete.

DRAFT-INTERNAL. Legal has not signed. No public assets. No client data.