social-health-tracker

name: social-health-tracker description: "Tracks social interactions, assesses loneliness via UCLA Loneliness Scale, maps Dunbar social circles, generates social health scores (0-100), and provides social prescriptions. Use when the user logs social activities, reports feeling isolated, or asks about social wellbeing patterns." version: 1.0.0 user-invocable: false allowed-tools: Read, Grep, Glob, Write, Edit metadata: {"openclaw":{"emoji":"🤝","category":"health"}}

Social Health Tracker

Tracks social interactions and assesses social wellbeing. Records social activities with quality ratings, administers the UCLA Loneliness Scale (short form), maps social connections using Dunbar's number framework, computes a Social Health Score (0-100), generates social prescriptions, analyzes social-health correlations, monitors boundary health, and detects isolation or over-extension patterns.

Capabilities

1. Social Activity Recording

Extract from natural-language input (e.g., "had coffee with Lisa for an hour", "family dinner tonight", "video call with college friends"):

Field	Required	Default if missing
date	Yes	today
time	No	omitted
duration (min)	No	60 min
type	Yes	see categories below
format	No	in-person
people	No	omitted
group_size	No	inferred from people, or 2
quality (1-5)	No	omitted (prompt user)
energy_change	No	omitted
notes	No	omitted

Interaction Types

Category	Examples
intimate	Deep 1-on-1 conversation, emotional support exchange, vulnerability sharing
social	Group dinner, party, casual hangout, coffee catch-up
professional	Work meeting, networking event, mentoring session
community	Volunteer work, religious gathering, club activity, neighborhood event
digital	Video call, voice call, meaningful text exchange, online gaming together
passive	Social media browsing, watching streams (low social value — flag if dominant)

Interaction Format

Format	Quality Multiplier	Notes
in-person	1.0×	Highest social value
video-call	0.8×	Good but lacks physical co-presence
voice-call	0.7×	Auditory connection only
text/chat	0.5×	Asynchronous, lower intimacy
social-media	0.3×	Passive; may increase loneliness if predominant

2. UCLA Loneliness Scale (Short Form, ULS-8)

Administer the 8-item version. Each item scored 1 (Never) to 4 (Always).

ULS-8 Items

I lack companionship
There is no one I can turn to
I am an outgoing person (reverse scored)
I feel left out
I feel isolated from others
I can find companionship when I want it (reverse scored)
I am unhappy being so withdrawn
People are around me but not with me

Scoring

Reverse score items 3 and 6: Never=4, Rarely=3, Sometimes=2, Always=1
Sum all 8 items → range 8-32

Score	Level	Interpretation
8-12	Low loneliness	Healthy social connectedness
13-18	Moderate loneliness	Some social needs unmet; review patterns
19-24	High loneliness	Significant isolation; social prescription recommended
25-32	Severe loneliness	Intervention needed; consider professional support

Frequency

Full assessment: Monthly (or when user requests)
Quick check: Single-item "How lonely have you felt this past week?" (1-10 scale) — weekly OK
Track trend: Score change > 4 points between assessments → flag

3. Social Health Score (0-100)

Score = Frequency(25%) + Quality(25%) + Diversity(20%) + Loneliness(20%) + Reciprocity(10%)

3.1 Frequency Sub-score (weight 25%)

Meaningful interactions per week (excluding passive/social-media):

Interactions/week	Points
≥ 7	100
5-6	85
3-4	65
1-2	40
0	10

3.2 Quality Sub-score (weight 25%)

Average quality rating (1-5) of interactions, adjusted by format multiplier:

Adjusted Quality = average(quality_i × format_multiplier_i) / 5 × 100

Adjusted Score	Points
≥ 80	100
60-79	80
40-59	55
20-39	30
< 20	10

3.3 Diversity Sub-score (weight 20%)

Number of distinct interaction categories (from the 6 types) used in the past 2 weeks:

Categories Active	Points
5-6	100
4	80
3	60
2	40
1	20
0	0

3.4 Loneliness Sub-score (weight 20%)

Inverted from UCLA score (lower loneliness = higher sub-score):

Loneliness Sub-score = max(0, (32 - UCLA_score) / 24 × 100)

UCLA Score	Sub-score
8	100
12	83
18	58
24	33
32	0

3.5 Reciprocity Sub-score (weight 10%)

Ratio of initiated vs. received social contacts over past 2 weeks:

Initiation Ratio	Points	Interpretation
40-60% (balanced)	100	Healthy reciprocity
30-39% or 61-70%	70	Slightly imbalanced
20-29% or 71-80%	40	Notable imbalance
< 20% or > 80%	15	One-sided pattern; review

Score Interpretation

Score	Level	Interpretation
90-100	Excellent	Thriving social life; maintain current patterns
75-89	Good	Healthy with minor gaps; review weakest dimension
60-74	Fair	Some social needs unmet; targeted improvements recommended
40-59	Poor	Significant social gaps; social prescription generated
< 40	Critical	Social isolation risk; professional support recommended

4. Dunbar Circle Framework

Map the user's social connections using Robin Dunbar's layered model:

Circle	Size	Bond Strength	Contact Frequency	Description
Support (Inner)	~5	Strongest	Weekly+	Closest confidants; call at 3 AM
Sympathy	~15	Strong	Monthly+	Close friends; would grieve their loss deeply
Affinity	~50	Moderate	Quarterly+	Good friends; regular social contact
Active	~150	Casual	Yearly+	Acquaintances; recognized and named

Circle Health Assessment

Indicator	Healthy	Warning
Support circle size	3-7 people	< 2 (isolation risk) or > 10 (diluted intimacy)
Sympathy circle	10-20 people	< 5 (narrow network)
Time invested in support circle	≥ 40% of social time	< 20% (neglecting core bonds)
Circle changes	Gradual evolution	Rapid loss of multiple support members
Cross-domain contacts	Friends from ≥ 3 life domains	Single-domain network (fragile)

5. Social Prescription

Generate personalized social prescriptions based on identified gaps.

Prescription Categories

Gap Identified	Prescription	Dose	Example
Low frequency	Increase contact	+2 interactions/week	Schedule 2 coffee dates
Low quality	Deepen existing bonds	1 intimate conversation/week	Ask a close friend a deep question
Low diversity	Expand categories	Add 1 new category/month	Join a community volunteer group
High loneliness	Structured social activity	3×/week minimum	Group class, regular meetup, community service
Low reciprocity (passive)	Initiate contact	2 outreach messages/week	Text a friend you haven't spoken to
Low reciprocity (over-giving)	Allow receiving	Practice accepting invitations	Say "yes" to next 3 invitations
Weak support circle	Deepen 1-2 connections	Weekly 1-on-1 time	Regular walk/meal with potential close friend
Digital-dominant	In-person shift	Replace 2 digital with in-person/week	Convert video call to coffee meeting

Prescription Format

Rx: [Action] — [Frequency] — [Duration]
Target: [Which sub-score this improves]
Start: [Date]
Review: [Date + 2 weeks]

6. Social-Health Correlation Analysis

Cross-reference social data with other health metrics (when available):

Correlation Pairs

Social Metric	Health Metric	Expected Relationship	Source
Weekly interactions	Sleep quality score	Positive (moderate)	Sleep Analyzer
Loneliness score	PHQ-9 depression	Positive (strong)	Mental Health Analyzer
Social diversity	Stress level	Negative (moderate)	Stress Management Coach
In-person time	Daily step count	Positive (weak-moderate)	Activity data
Social quality avg	Mood score	Positive (strong)	Mood Tracker
Isolation days (0 contacts)	Anxiety (GAD-7)	Positive (moderate)	Mental Health Analyzer

Analysis Rules

Require ≥ 14 days of overlapping data before reporting correlations
Use Pearson correlation coefficient; report only |r| ≥ 0.3
Always note: "Correlation does not imply causation"
Present as trend observation, not medical finding

7. Boundary Health

Monitor for unhealthy social patterns at both extremes:

Over-Extension Indicators

Signal	Threshold	Response
Social hours/week	> 25h and energy_change frequently negative	Flag over-extension
Quality dropping with quantity rising	Quality avg < 3 while frequency > 10/week	Suggest reducing low-quality interactions
Cancellation rate	> 30% of planned events cancelled	Possible burnout; reduce commitments
Recovery time needed	Reports needing ≥ 1 full day alone after socializing	Respect introversion; optimize not maximize

Withdrawal Indicators

Signal	Threshold	Response
Zero-interaction days	≥ 3 consecutive days	Gentle prompt: "Haven't logged social activity in 3 days — everything OK?"
Declining invitations	> 50% declined in 2 weeks	Check for avoidance patterns
Quality all low	Average < 2 for past week	May indicate depression; cross-check mood
Only passive social	> 80% of logged interactions are social-media type	Flag: passive social may increase loneliness

8. Pattern Detection

Weekly Patterns

Social rhythm: Identify preferred social days/times
Energy mapping: When does socializing energize vs. drain?
Weekend vs. weekday: Balance check

Monthly Trends

Score trajectory (improving, stable, declining)
Circle changes (people entering/leaving inner circles)
Seasonal patterns (e.g., winter isolation, holiday over-extension)

Critical Alerts

Pattern	Trigger	Action
Rapid isolation	Score dropped > 20 points in 2 weeks	Flag concern; suggest reaching out to support circle
Chronic loneliness	UCLA ≥ 19 for 2+ consecutive assessments	Recommend professional support (therapist, counselor)
Network collapse	Lost ≥ 2 support circle members in 1 month	Check in; life transition support
Social anxiety pattern	High avoidance + high loneliness	Suggest graduated exposure; professional referral

Output Format

Social Activity Log

## 🤝 Social Activity Logged

### Entry
- **Date**: 2026-03-10, 19:00-21:00
- **Type**: Social (group dinner)
- **Format**: In-person (1.0×)
- **People**: Lisa, Mark, Jenny
- **Group size**: 4
- **Quality**: 4/5
- **Energy change**: +1 (energized)
- **Duration**: 120 min

### Running Weekly Stats
- Interactions this week: 5 (target: ≥ 5 ✅)
- Average quality: 3.8/5
- In-person ratio: 60%
- Categories active: 3/6 (social, professional, digital)

Weekly Social Report

## 🤝 Weekly Social Report (MM-DD ~ MM-DD)

### Social Health Score: 71 / 100
| Dimension | Raw | Sub-score | Weight | Weighted |
|-----------|-----|-----------|--------|----------|
| Frequency | 5/week | 85 | 25% | 21.3 |
| Quality | 3.6 adj | 72 | 25% | 18.0 |
| Diversity | 3 categories | 60 | 20% | 12.0 |
| Loneliness | UCLA 15 | 71 | 20% | 14.2 |
| Reciprocity | 55% initiated | 100 | 10% | 10.0 |
| **Total** | | | | **75.5** |

### Interaction Summary
| Day | Interactions | Best Moment | Energy |
|-----|-------------|------------|--------|
| Mon | 1 (work meeting) | -- | Neutral |
| Tue | 0 | -- | -- |
| Wed | 2 (coffee + call) | Deep talk with Alex | +1 |
| Thu | 1 (team lunch) | -- | Neutral |
| Fri | 1 (dinner with friends) | Group laughter | +2 |
| Sat | 0 | -- | -- |
| Sun | 1 (family video call) | Catching up with parents | +1 |

### Dunbar Circle Activity
| Circle | People Contacted | Expected Minimum | Status |
|--------|-----------------|-----------------|--------|
| Support (~5) | 3/5 | Weekly | ⚠️ 2 not contacted |
| Sympathy (~15) | 4/12 | Monthly | ✅ On track |

### Patterns
- [Good] Reciprocity is balanced (55% initiated)
- [Notice] 2 zero-interaction days; both weekend — consider light social plans
- [Notice] Only 3/6 interaction categories active — try community or hobby activity
- [Alert] Support circle member "Mom" not contacted in 2 weeks

### Social Prescription
Rx: Schedule 1 community activity — Weekly — Start this week
Target: Diversity sub-score (currently 60 → target 80)
Review: MM-DD

Loneliness Assessment

## 🤝 UCLA Loneliness Assessment

### ULS-8 Results
| # | Item | Response | Score |
|---|------|----------|-------|
| 1 | I lack companionship | Sometimes | 3 |
| 2 | There is no one I can turn to | Rarely | 2 |
| 3 | I am an outgoing person | Sometimes | 3 (R) |
| 4 | I feel left out | Sometimes | 3 |
| 5 | I feel isolated from others | Rarely | 2 |
| 6 | I can find companionship when I want | Sometimes | 3 (R) |
| 7 | I am unhappy being so withdrawn | Rarely | 2 |
| 8 | People are around me but not with me | Sometimes | 3 |
| | **Total** | | **21** |

### Interpretation
- **Level**: High loneliness (score 21/32)
- **Previous**: 17 (moderate) on MM-DD → **increased +4 points** ⚠️
- **Key items**: "Lack companionship" and "people around but not with me" scored highest

### Recommendations
- Social prescription intensity: High
- Focus: Quality of existing interactions over quantity
- Suggested: 1 intimate conversation/week + 1 new group activity
- Consider: Speaking with a counselor if score remains ≥ 19 for 2+ months

Data Persistence

Daily File (`memory/health/daily/YYYY-MM-DD.md`)

## Social [social-health-tracker · HH:MM]
- Interactions: 2 (coffee with Lisa, video call with parents)
- Quality avg: 4.0/5
- Types: social, digital
- Format: 1 in-person, 1 video-call
- Energy: +1 overall
- Weekly running score: 71/100

Item File (`memory/health/items/social-health.md`)

---
item: social-health
unit: score (0-100)
updated_at: YYYY-MM-DDTHH:MM:SS
---

# Social Health Records

## Profile
- Personality tendency: Ambivert (estimated)
- Optimal interactions/week: 5-7
- Preferred formats: In-person, video call
- Dunbar support circle: Alex, Mom, Dad, Lisa, Sam

## Recent Status
- Latest: Score 71, 5 interactions/week (YYYY-MM-DD)
- 4-week average score: 68
- UCLA loneliness (latest): 15 (moderate)
- Trend: Improving

## UCLA History
| Date | Score | Level | Change |
|------|-------|-------|--------|
| 2026-03-10 | 15 | Moderate | -2 |
| 2026-02-10 | 17 | Moderate | -- |

## Weekly History
| Week | Score | Interactions | Quality Avg | Diversity | UCLA | Notes |
|------|-------|-------------|------------|-----------|------|-------|
| 03-04~03-10 | 71 | 5 | 3.8 | 3/6 | 15 | |
| 02-25~03-03 | 65 | 3 | 3.5 | 2/6 | 17 | Low week |

Write Steps

Glob memory/health/daily/ for the target date file; create if absent.
Read the daily file.
Grep ## .* \[social-health-tracker · for existing section.
Edit to replace existing section, or insert before ## Health Files if new.
Update frontmatter updated_at.
Read memory/health/items/social-health.md (create if absent).
Update ## Recent Status and prepend new row to ## Weekly History.
Update ## UCLA History if a new assessment was completed.
Remove rows older than 6 months.

Alerts and Safety

Medical Disclaimer

This skill is for social wellbeing tracking only and does not constitute medical or psychological advice. It does not diagnose social anxiety disorder, depression, or other mental health conditions.

Consult a professional if:

UCLA loneliness score ≥ 19 persists for 2+ months
Social withdrawal is accompanied by depressive symptoms (PHQ-9 ≥ 10)
Social anxiety significantly limits daily functioning (avoiding work, school, errands)
Grief or loss of relationship causing prolonged dysfunction (> 6 months)
Social isolation is accompanied by substance use as coping mechanism

Crisis Awareness

Social isolation is a significant risk factor for mental health crises.

If the user expresses:

"Nobody cares about me" / "I have no one"
"The world would be better without me"
"I don't want to be here anymore"

→ Immediately display crisis resources (see mental-wellness-companion) and pause social tracking.

Ethical Boundaries

Never judge the user's social preferences or introversion as pathological
Respect that optimal social levels vary greatly between individuals
Do not pressure the user to socialize more if they express satisfaction with lower levels
Acknowledge cultural differences in social norms and expectations
Protect privacy: Never share names or relationship details outside the user's health data files
Avoid labeling: Use "pattern observed" language, not diagnostic labels

Introversion vs. Isolation

Important distinction:

Healthy introversion: Lower social frequency by choice, high quality, low loneliness → healthy; do not pathologize
Isolation: Low frequency, low quality, high loneliness, declining function → concerning; offer support

Calibrate recommendations to the user's baseline, not population norms. An introvert scoring 3 interactions/week with high quality and low loneliness has excellent social health.