Sleep analyzer

Sleep Analyzer Skill

Analyze sleep data, identify sleep patterns, assess sleep quality, and provide personalized sleep-improvement advice.

Functions

1. Sleep-trend analysis

Analyze trends in sleep duration, quality, and efficiency; identify what is improving or needs attention.

Analysis dimensions:

• Sleep-duration trend (change in average duration)
• Sleep-efficiency trend (change in efficiency %)
• Sleep-timing pattern (bedtime, sleep-onset, wake time)
• Sleep-consistency score
• Weekend vs. weekday comparison (social jet lag)

Output:

• Trend direction (improving/stable/declining)
• Magnitude and percentage of change
• Trend-significance assessment
• Best sleep-window identification
• Improvement suggestions

2. Sleep-quality assessment

Comprehensively assess sleep quality and identify key factors affecting it.

Assessment content:

• PSQI score tracking and trend
• Subjective sleep-quality distribution (good/fair/poor)
• Nighttime-awakening analysis (count, duration, cause)
• Sleep-stage analysis (deep, light, REM proportions)
• Post-sleep restfulness assessment

Output:

• Sleep-quality grade (excellent/good/fair/poor)
• Quality trend
• Main influencing factors
• Quality-improvement priorities

3. Sleep-problem identification

Identify common sleep problems and risk factors.

Identification content:

• Insomnia patterns:

  • Difficulty falling asleep (sleep latency >30 min)
  • Difficulty staying asleep (>2 awakenings or total wake time >30 min)
  • Early awakening (waking >30 min earlier than intended)
  • Mixed insomnia

• Sleep-apnea risk:

  • STOP-BANG questionnaire score
  • Symptom analysis (snoring, gasping awake, daytime sleepiness)
  • Risk level (low/medium/high)

• Other issues:

  • Irregular-schedule detection
  • Sleep-debt calculation (ideal vs. actual duration)
  • Social-jet-lag assessment

Output:

• Whether a problem is present
• Problem type and severity
• Risk-factor list
• Whether to recommend seeking care

4. Correlation analysis

Analyze the correlation between sleep and other health metrics.

Supported correlations:

• Sleep ↔ exercise:

  • Sleep difference on exercise vs. rest days
  • Effect of exercise timing (morning/afternoon/evening)
  • Exercise intensity vs. sleep quality

• Sleep ↔ diet:

  • Caffeine intake vs. sleep duration and onset
  • Alcohol's effect on sleep structure
  • Dinner time vs. sleep quality

• Sleep ↔ mood:

  • Bidirectional sleep–mood relationship
  • Stress level vs. sleep quality
  • Sleep deprivation vs. daytime mood

• Sleep ↔ chronic disease:

  • Sleep vs. hypertension
  • Sleep vs. glycemic control
  • Sleep vs. weight change

Output:

• Correlation coefficient (-1 to 1)
• Correlation strength (weak/medium/strong)
• Statistical significance
• Causal-relationship inference
• Practical suggestions

5. Personalized recommendations

Generate personalized sleep-improvement advice from the user's data.

Recommendation types:

• Schedule adjustment:

  • Best bedtime/wake time
  • Consistency-improvement plan
  • Nap-management advice

• Pre-sleep preparation:

  • Bedtime-routine design
  • Relaxation-technique recommendations
  • Screen-time management

• Sleep-environment optimization:

  • Temperature, humidity, light, noise
  • Bedding-comfort suggestions

• Lifestyle adjustments:

  • Exercise, diet, caffeine, alcohol management
  • Stress-management advice

• CBT-I elements:

  • Stimulus control
  • Sleep restriction
  • Cognitive restructuring

Output:

• Priority-ranked suggestion list
• Specific implementation steps
• Expected effects
• Implementation timeline

---

Usage

Triggers

Trigger this skill when the user requests:

• Sleep-trend analysis
• Sleep-quality assessment
• Sleep-problem identification
• Sleep-improvement advice
• Correlation analysis between sleep and other health metrics

Execution Steps

Step 1: Determine the analysis scope

Clarify the requested analysis type and time range:

• Type: trend/quality/problems/correlation/advice
• Time range: week/month/quarter/custom

Step 2: Read data

Primary data sources:

1. data-example/sleep-tracker.json - main sleep data
2. data-example/sleep-logs/YYYY-MM/YYYY-MM-DD.json - daily sleep records

Related data sources:

1. data-example/fitness-tracker.json - exercise data
2. data-example/hypertension-tracker.json - blood-pressure data
3. data-example/diabetes-tracker.json - blood-glucose data
4. data-example/diet-records/ - food records
5. data-example/mood-tracker.json - mood data

Step 3: Data analysis

Run the appropriate algorithm for the analysis type:

Trend analysis:

• Linear regression for trend slope
• Moving average to smooth fluctuation
• Statistical-significance test

Correlation analysis:

• Pearson correlation
• Lagged-correlation analysis (delayed effects)
• Multivariate regression

Pattern recognition:

• Time-series pattern recognition
• Outlier detection
• Periodicity analysis

Step 4: Generate the report

Output the analysis report in the standard format (see "Output Format").

---

Output Format

Sleep-quality analysis report

# Sleep Quality Analysis Report

## Analysis period
2025-03-20 to 2025-06-20 (3 months)

---

## Sleep-duration trend

- **Trend**: ⬆️ improving
- **Start**: avg 6.2 hours/night
- **Current**: avg 7.1 hours/night
- **Change**: +0.9 hours (+14.5%)
- **Interpretation**: marked increase, near the ideal target (7.5 hours)

---

## Sleep efficiency

- **Average sleep efficiency**: 85.3%
- **Range**: 78%-92%
- **On-target rate**: 63% (>85% is on target)
- **Interpretation**: normal efficiency with room to improve

**Efficiency distribution**:
- Excellent (>90%): 15 nights
- Good (85-90%): 28 nights
- Needs improvement (<85%): 47 nights

---

## Schedule regularity

- **Average bedtime**: 23:15 (range: 22:30-01:00)
- **Average wake time**: 07:05 (range: 06:30-08:30)
- **Consistency score**: 72/100
- **Social jet lag**: 45 minutes (later sleep/wake on weekends)
- **Interpretation**: largely regular, but with large weekend variability

**Suggestions**:
- 🎯 Keep a consistent wake time, including weekends
- 🎯 Adjust bedtime gradually; avoid excessive weekend delay

---

## Sleep-quality distribution

| Quality grade | Nights | Share | Trend |
|---------------|--------|-------|-------|
| Excellent | 8 | 9% | ⬆️ |
| Very good | 12 | 13% | ➡️ |
| Good | 15 | 17% | ⬆️ |
| Fair | 42 | 47% | ⬇️ |
| Poor | 10 | 11% | ⬇️ |
| Very poor | 3 | 3% | ➡️ |

**Interpretation**: quality is mostly "fair", but nights of "good"+ are increasing

---

## Nighttime-awakening analysis

- **Average awakenings**: 1.8/night
- **Average awakening duration**: 18 minutes
- **Main causes**:
  1. Need to urinate (45%)
  2. Noise (25%)
  3. Too warm (15%)
  4. Other (15%)

**Suggestions**:
- 🎯 Limit fluids for 2 hours before bed
- 🎯 Optimize bedroom temperature (18-22℃)
- 🎯 Use a white-noise machine to mask background noise

---

## PSQI assessment trend

- **Latest score**: 8 (fair sleep quality)
- **Previous score**: 10 (2025-03-20)
- **Change**: -2 (improvement)
- **Trend**: ⬆️ continued improvement

**Component changes**:
- Subjective sleep quality: 2→2 (stable)
- Sleep latency: 2→2 (stable)
- Sleep duration: 2→1 (improving)
- Sleep efficiency: 2→1 (improving)
- Sleep disturbance: 2→1 (improving)

---

## Sleep-problem identification

### Insomnia assessment

- **Type**: mixed insomnia
- **Frequency**: 4-5 nights/week
- **Duration**: 18 months
- **Main symptoms**:
  - ✗ Difficulty falling asleep (latency >30 min)
  - ✗ Difficulty staying asleep (>2 awakenings)
  - ✓ No early awakening

- **Impact**:
  - Daytime fatigue: moderate
  - Irritability: yes
  - Concentration difficulty: yes
  - Work performance: mildly affected

- **Suggestion**: 🏥 lasting >3 months; recommend consulting a sleep specialist

### Sleep-apnea screening (STOP-BANG)

- **Score**: 3/8
- **Risk level**: medium
- **Positive items**:
  - ✗ Snoring
  - ✗ Tired (daytime fatigue)
  - ✓ Observed apnea (none observed)
  - ✗ Pressure (hypertension)
  - ✓ BMI > 28
  - ✓ Age > 50
  - ✗ Neck size > 40cm
  - ✓ Gender = male

- **Suggestion**: ⚠️ recommend a sleep study (PSG)

---

## Correlation analysis

### Sleep ↔ exercise

**Exercise day vs. rest day**:
- Exercise-day average sleep: 7.3 hours
- Rest-day average sleep: 6.8 hours
- Difference: +0.5 hours (+7.4%)

**Effect of exercise timing**:
- Morning exercise: sleep 7.5 h, quality 7.8/10
- Afternoon exercise: sleep 7.2 h, quality 7.5/10
- Evening exercise: sleep 6.8 h, quality 6.8/10

**Correlation**: medium positive (r = 0.42)
**Conclusion**: regular exercise improves sleep, but avoid vigorous exercise 2-3 hours before bed

**Suggestions**:
- 🎯 Keep a regular exercise habit
- 🎯 Move exercise to morning or afternoon
- 🎯 Avoid vigorous exercise 2-3 hours before bed

---

### Sleep ↔ caffeine

**Caffeine-timing analysis**:
- Intake before 2 PM: avg sleep 7.2 h, latency 25 min
- Intake after 2 PM: avg sleep 6.7 h, latency 40 min
- Difference: -0.5 h duration, +15 min latency

**Correlation**: medium negative (r = -0.38)
**Conclusion**: caffeine after 2 PM significantly affects sleep

**Suggestions**:
- 🎯 Avoid caffeine after 2 PM
- 🎯 Avoid caffeine entirely 6 hours before bed

---

### Sleep ↔ mood

**Effect of sleep quality on next-day mood**:
- Good sleep: 82% chance of positive next-day mood
- Fair sleep: 45% chance
- Poor sleep: 18% chance

**Effect of pre-sleep mood on falling asleep**:
- High pre-sleep stress: latency 45 min
- Low pre-sleep stress: latency 20 min
- Difference: +25 min

**Correlation**: strong bidirectional (r = 0.65)
**Conclusion**: sleep and mood significantly affect each other

**Suggestions**:
- 🎯 Do stress management before bed (meditation, deep breathing)
- 🎯 Build a relaxing bedtime routine
- 🎯 Keep a mood journal to identify stress patterns

---

## Insights and Suggestions

### Key insights

1. **Schedule inconsistency is the main problem**
   - Social jet lag of 45 min
   - Weekend schedule deviates markedly from weekdays
   - Affects: circadian disruption, "Monday jet lag"

2. **Evening exercise affects sleep onset**
   - Latency +15 min on evening-exercise days
   - Suggestion: adjust exercise timing

3. **Sleep environment can be optimized**
   - Noise accounts for 25% of awakenings
   - Too warm accounts for 15%
   - Suggestion: targeted improvements

---

### Priority action plan

#### Priority 1: build a consistent schedule (2 weeks)

**Goal**: raise consistency score to 85

**Specific actions**:
1. Fixed wake time 07:00 (including weekends)
2. Fixed bedtime 23:00
3. Limit naps to <30 min, before 3 PM
4. Adjust weekend schedule gradually (15 min earlier each time)

**Expected effect**:
- Consistency score: 72 → 85
- Efficiency: +3-5%
- Less Monday fatigue

---

#### Priority 2: create a bedtime routine (3 weeks)

**Goal**: establish a stable bedtime routine

**Specific actions**:
1. Start the routine 1 hour earlier (22:00)
2. Turn off electronics (22:30)
3. Dim bedroom lighting
4. Relaxing activity (reading, meditation, warm bath)
5. Keep the bedroom quiet, dark, and cool (18-22℃)

**Expected effect**:
- Latency: 30 → 20 min
- Quality: fair → good
- Lower pre-sleep stress

---

#### Priority 3: optimize the sleep environment (1 week)

**Goal**: remove environmental disturbances

**Specific actions**:
1. Install blackout curtains
2. Use a white-noise machine
3. Optimize temperature to 18-22℃
4. Remove the bedroom clock
5. Replace pillow and mattress for comfort

**Expected effect**:
- Awakenings: 1.8 → 1.2/night
- Better sleep continuity
- Better morning state

---

#### Priority 4: lifestyle adjustments (4 weeks)

**Goal**: remove habits that harm sleep

**Specific actions**:
1. Move exercise to morning or afternoon
2. Stop caffeine after 2 PM
3. Avoid alcohol 3 hours before bed
4. Avoid large meals 2 hours before bed
5. Avoid work-related discussion 1 hour before bed

**Expected effect**:
- Duration: +0.3 h
- Quality: +1 point
- PSQI: 8 → 6

---

## Long-term goals

- **Sleep duration**: reach 7.5 hours/night (currently 7.1)
- **Sleep efficiency**: >90% (currently 85%)
- **PSQI score**: ≤5 (currently 8)
- **Consistency**: ≥85 (currently 72)
- **Sleep latency**: <20 min (currently 28)

---

## Medical Safety Reminder

⚠️ **Care-seeking advice**:
- 🏥 Insomnia lasting >3 months: consult a sleep specialist
- 🏥 STOP-BANG ≥3: recommend a sleep study (PSG)
- 🏥 Severe sleepiness affecting driving safety: seek care immediately

---

**Report generated**: 2025-06-20
**Analysis period**: 2025-03-20 to 2025-06-20 (90 days)
**Records**: 90 nights
**Sleep Analyzer version**: v1.0

---

Data Structure

Sleep-record data

{
  "sleep_records": [
    {
      "id": "sleep_20250620001",
      "date": "2025-06-20",
      "sleep_times": {
        "bedtime": "23:00",
        "sleep_onset_time": "23:30",
        "wake_time": "07:00",
        "out_of_bed_time": "07:15"
      },
      "sleep_metrics": {
        "sleep_duration_hours": 7.0,
        "time_in_bed_hours": 8.25,
        "sleep_latency_minutes": 30,
        "sleep_efficiency": 84.8
      },
      "sleep_quality": {
        "subjective_quality": "fair",
        "quality_score": 5,
        "rested_feeling": "somewhat"
      },
      "factors": {
        "exercise": true,
        "exercise_time": "evening",
        "caffeine_after_2pm": false,
        "screen_time_before_bed_minutes": 60
      }
    }
  ]
}

---

Algorithm Notes

Sleep-quality scoring

def calculate_sleep_quality_score(record):
    """
    Compute a sleep-quality score (0-10)

    Factor weights:
    - Duration: 30%
    - Efficiency: 25%
    - Sleep latency: 20%
    - Nighttime awakenings: 15%
    - Subjective quality: 10%
    """
    score = 0

    # Duration score (ideal 7-9 hours)
    duration = record['sleep_duration_hours']
    if 7 <= duration <= 9:
        duration_score = 10
    elif 6 <= duration < 7 or 9 < duration <= 10:
        duration_score = 7
    else:
        duration_score = 4
    score += duration_score * 0.30

    # Efficiency score (>90% excellent)
    efficiency = record['sleep_efficiency']
    efficiency_score = min(efficiency / 90 * 10, 10)
    score += efficiency_score * 0.25

    # Latency score (<15 min excellent)
    latency = record['sleep_latency_minutes']
    if latency <= 15:
        latency_score = 10
    elif latency <= 30:
        latency_score = 7
    elif latency <= 45:
        latency_score = 4
    else:
        latency_score = 1
    score += latency_score * 0.20

    # Awakenings score (0 excellent)
    awakenings = record['awakenings']['count']
    awakening_score = max(10 - awakenings * 2, 0)
    score += awakening_score * 0.15

    # Subjective-quality score
    quality_map = {
        'excellent': 10,
        'very_good': 8,
        'good': 7,
        'fair': 5,
        'poor': 3,
        'very_poor': 1
    }
    subjective_score = quality_map.get(
        record['sleep_quality']['subjective_quality'],
        5
    )
    score += subjective_score * 0.10

    return round(score, 1)

Schedule-regularity scoring

def calculate_sleep_consistency_score(records):
    """
    Compute a schedule-regularity score (0-100)

    Factors:
    - Bedtime standard deviation
    - Wake-time standard deviation
    - Duration standard deviation
    - Weekday vs. weekend difference
    """
    # Extract time data
    bedtimes = [r['bedtime'] for r in records]
    wake_times = [r['wake_time'] for r in records]
    durations = [r['sleep_duration_hours'] for r in records]

    # Standard deviation (minutes)
    bedtime_std = time_to_minutes_std(bedtimes)
    wake_std = time_to_minutes_std(wake_times)
    duration_std = statistics.stdev(durations)

    # Weekday vs. weekend difference
    weekday_avg = avg([r['sleep_duration_hours']
                       for r in records if is_weekday(r)])
    weekend_avg = avg([r['sleep_duration_hours']
                       for r in records if is_weekend(r)])
    diff = abs(weekday_avg - weekend_avg)

    # Overall score
    score = 100
    score -= bedtime_std * 0.5  # bedtime SD effect
    score -= wake_std * 0.5     # wake-time SD effect
    score -= duration_std * 2   # duration SD effect
    score -= diff * 10          # weekday/weekend difference

    return max(0, min(100, round(score)))

Correlation analysis

def calculate_correlation(sleep_data, other_data, lag_days=0):
    """
    Compute the correlation between sleep and another metric

    Args:
    - sleep_data: list of sleep data
    - other_data: list of other-metric data
    - lag_days: lag in days (for delayed effects)

    Returns:
    - correlation_coefficient
    - p_value
    - interpretation
    """
    # Align data (accounting for lag)
    aligned = align_data_with_lag(sleep_data, other_data, lag_days)

    # Pearson correlation
    from scipy import stats
    corr, p_value = stats.pearsonr(
        aligned['sleep_values'],
        aligned['other_values']
    )

    # Interpret
    if abs(corr) < 0.3:
        strength = "weak"
    elif abs(corr) < 0.7:
        strength = "medium"
    else:
        strength = "strong"

    direction = "positive" if corr > 0 else "negative"
    significant = p_value < 0.05

    interpretation = f"{strength} {direction} correlation"
    if significant:
        interpretation += " (statistically significant)"

    return {
        'correlation_coefficient': round(corr, 3),
        'p_value': round(p_value, 4),
        'interpretation': interpretation,
        'significant': significant
    }

---

Medical Safety Statement

The analysis and advice provided by this skill is for reference only and does not constitute a medical diagnosis or treatment plan.

What this skill CAN do:

• ✅ Analyze sleep data and patterns
• ✅ Identify sleep-problem risk
• ✅ Provide sleep-hygiene advice
• ✅ Assess correlation with other health metrics

What this skill CANNOT do:

• ❌ Diagnose insomnia, sleep apnea, etc.
• ❌ Prescribe sleep aids or treatment
• ❌ Replace professional sleep-medicine care
• ❌ Manage severe sleep disorders

When to seek care:

• 🏥 Insomnia lasting >3 months
• 🏥 Suspected sleep apnea (STOP-BANG ≥3)
• 🏥 Severe sleepiness affecting safety
• 🏥 Sudden severe sleep problems

---

Reference Resources

• AASM sleep-scoring standards: https://aasm.org/
• PSQI scale: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3455216/
• STOP-BANG questionnaire: https://www.stopbang.ca/
• CBT-I therapy: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3455216/

---

Skill version: v1.0 Created: 2026-01-02 Maintainer: MedClawMini