skills/nirholas/xactions/community-health-monitoring

community-health-monitoring

SKILL.md

Community Health Monitoring

MCP-powered workflow for auditing follower quality, engagement health, and network efficiency. Produces a scored health report.

MCP Tools Used

Tool Purpose
x_get_profile Account-level stats
x_get_followers Follower list for quality audit
x_get_following Following list for reciprocity check
x_get_non_followers Identify non-reciprocal follows
x_get_tweets Engagement data for authenticity check
x_detect_unfollowers Track recent unfollower patterns

Browser Scripts

Complement MCP analysis with browser-side tools:

Goal Script
Audit follower quality src/auditFollowers.js
Detect unfollowers src/detectUnfollowers.js
Audience demographics src/audienceDemographics.js
Follow ratio analysis src/followRatioManager.js
Account health dashboard src/accountHealthMonitor.js
Shadowban check src/shadowbanChecker.js

Workflow

  1. Profile baseline -- Call x_get_profile to get follower count, following count, and calculate follower-to-following ratio.
  2. Audit follower quality -- Call x_get_followers with limit: 200. Classify each follower:
    • Active: Has bio, 50+ followers, posted in last 30 days
    • Low quality: No bio, <10 followers, or no recent activity
    • Suspect bot: Default avatar, username with many numbers, 0 tweets, follows 1000+
  3. Check engagement authenticity -- Call x_get_tweets with limit: 30. For each tweet, compare engagement volume to follower count. Flag anomalies: likes/follower ratio > 10% (potential engagement pods) or < 0.1% (ghost followers).
  4. Analyze unfollower patterns -- Call x_detect_unfollowers. Note churn rate and whether unfollowers correlate with specific content types or posting gaps.
  5. Assess reciprocity -- Call x_get_non_followers. Calculate reciprocity rate: mutual_follows / total_following * 100. Identify high-value accounts not following back.
  6. Calculate health score -- Weighted composite (0-100):
    • Follower quality: 30% (% active followers)
    • Engagement authenticity: 25% (normal engagement patterns)
    • Churn rate: 20% (low unfollower rate)
    • Reciprocity: 15% (healthy follower/following balance)
    • Growth trend: 10% (net positive follower change)
  7. Generate report -- Compile into the template below with actionable recommendations.

Output Template

## Community Health Report: @{username}
Date: {date} | Health Score: {score}/100

### Score Breakdown
| Category | Score | Weight | Weighted |
|----------|-------|--------|----------|
| Follower Quality | {n}/100 | 30% | {n} |
| Engagement Authenticity | {n}/100 | 25% | {n} |
| Churn Rate | {n}/100 | 20% | {n} |
| Reciprocity | {n}/100 | 15% | {n} |
| Growth Trend | {n}/100 | 10% | {n} |

### Follower Audit
- Total: {count} | Active: {n}% | Low quality: {n}% | Suspect bots: {n}%

### Engagement Health
- Avg engagement rate: {rate}%
- Anomalous posts: {count} flagged

### Reciprocity
- Following: {count} | Follow back: {n}% | Non-followers: {count}

### Recommendations
1. {actionable recommendation}
2. {actionable recommendation}
3. {actionable recommendation}

Strategy Guide

Monthly health audit routine

  1. Run full MCP workflow above for baseline report
  2. Compare against previous month's scores
  3. Action items: block flagged bots, unfollow non-reciprocals above threshold
  4. Use src/accountHealthMonitor.js for quick between-audit checks

Score interpretation

Score Grade Action
80-100 Excellent Maintain current strategy
60-79 Good Minor adjustments needed
40-59 Fair Review engagement strategy, clean follower list
20-39 Poor Major cleanup needed, block bots, reassess content
0-19 Critical Possible shadowban, mass bot followers, or inactive account

Improving a low health score

  1. Block suspect bot followers with src/blockBots.js
  2. Unfollow non-reciprocals with src/unfollowback.js
  3. Increase posting consistency to reduce churn
  4. Engage authentically to improve engagement rate

Notes

  • Health score is a heuristic -- use as directional guidance, not exact measurement
  • Bot detection uses profile signals, not ML -- some false positives expected
  • Run quarterly for trend tracking, monthly for active management
Weekly Installs
4
GitHub Stars
108
First Seen
Feb 28, 2026
Installed on
openclaw4
gemini-cli4
github-copilot4
codex4
kimi-cli4
cursor4