---
name: two-week-sleep-audit
description: "Runs a structured sleep audit for persistent tiredness: checks your sleep window, picks the best-evidenced factors and designs a two-week one-variable experiment. Use when a chronically tired person wants a structured review of their sleep schedule and habits that ends in a two-week experiment. Do not use for diagnosing a sleep disorder or recommending sleep medication. Related topics: Sleep, Health, Wellness."
---

# Two-Week Sleep Audit

Runs a structured sleep audit for persistent tiredness: checks your sleep window, picks the best-evidenced factors and designs a two-week one-variable experiment.

## When to use

Use this skill when a chronically tired person wants a structured review of their sleep schedule and habits that ends in a two-week experiment. Do not use it for diagnosing a sleep disorder or recommending sleep medication.

## Workflow

1. Ask the user for each input listed below in a single message, skipping any they have already given. Where options are listed, offer them, but accept other answers too.
2. Compute the actual sleep opportunity from bedtime, sleep onset and alarm and compare it with typical adult need.
3. Choose the one or two factors in the answers with the strongest evidence and name popular fixes with weak evidence to skip.
4. Design a two-week experiment that changes one variable and defines a one-minute morning log.
5. State what result would confirm or rule out the hypothesis and list concrete signs that warrant seeing a doctor.

## Inputs

- **Weeknight bedtime** (free text, e.g. "time in bed")
- **Asleep by roughly** (free text, e.g. "time")
- **Alarm** (free text, e.g. "time")
- **Weekend drift** (choose one: Basically the same, 1 to 2 hours later, 3 or more hours later, Totally different schedule)
- **Caffeine** (free text, e.g. "What and how much and the last one of the day")
- **What wakes me at night** (choose any: Nothing that I know of, Bathroom, Partner or kids, Noise or light, Racing thoughts, Wake up and do not know why)
- **Screens in bed** (choose one: Yes, No)
- **Anything else relevant** (longer free text)

## Output

A sleep opportunity calculation, the one or two best-evidenced factors with weak fixes to skip, a two-week one-variable experiment with its daily log, and a concrete list of doctor-worthy signs.

## Prompt template

Fill the inputs into this template and follow it as the full instruction.

### Two-Week Sleep Audit

Act as a sleep-focused health coach running me through a proper audit. I am tired most days and generic sleep hygiene advice has not helped.

My situation:
- Weeknight bedtime: {{Weeknight bedtime}}
- Asleep by roughly: {{Asleep by roughly}}
- Alarm: {{Alarm}}
- Weekend drift: {{Weekend drift}}
- Caffeine: {{Caffeine}}
- What wakes me at night: {{What wakes me at night}}
- Screens in bed: {{Screens in bed}}
- Anything else relevant: {{Anything else relevant}}

Work in this order. First, compare my actual sleep opportunity (bedtime to alarm, minus time to fall asleep) with what an adult usually needs, because if that arithmetic fails nothing else matters. Second, pick the one or two factors in my answers with the strongest evidence behind them, and say which popular fixes have weak evidence so I do not spend effort there. Third, design a two-week experiment: change one variable, say what to log each morning in under a minute, and state what result would confirm or rule out the hypothesis. Finish with concrete signs (loud snoring with pauses, falling asleep while driving, mood change) that mean it is beyond self-help and worth raising with a doctor.

## Pitfalls

- Changing caffeine, screens and bedtime together so the experiment proves nothing.
- Skipping the sleep-window arithmetic when the schedule leaves under seven hours.
- Ending with a boilerplate disclaimer instead of specific red-flag symptoms.
