cleanexpo Pi-SEO Health Monitor Skill
Analyse health trends across all monitored repos. Detect regressions, sustained degradation, and systemic cross-repo patterns. Generate portfolio health digests for operator review.
Data Sources
All data is read from .harness/scan-results/{project-id}/ — JSON files named {date}-{scan_type}.json.
Load the two most recent files per scan type per project for delta computation. Load up to 10 files per project for trend analysis.
Portfolio Health Score
Weighted average across all projects, where weight = scan_priority from projects.json:
| scan_priority | Weight |
|---|---|
| high | 3 |
| medium | 2 |
| low | 1 |
portfolio_health = sum(project.health_score * weight) / sum(weight)
Each project's health_score is the overall score from its most recent scan result (0-100, already computed by ScanResult.health_score).
Regression Detection
Compare each project's current overall score against the previous scan cycle:
- Regression: score dropped ≥ 5 points between consecutive scans
- Critical regression: score dropped ≥ 15 points
- Sustained degradation: overall score below 70 for 3+ consecutive scans
Track per scan type as well as the overall composite.
Systemic Issue Detection
A finding is systemic if the same scan_type + title combination appears in 2+ distinct repos within the same scan cycle. Systemic issues indicate a shared dependency, shared code pattern, or shared configuration problem.
Group findings by scan_type:title key across all current scan results. Any key with matches in ≥ 2 project IDs is systemic.
Alert Thresholds
Generate an alert for each of the following:
| Condition | Alert type | Severity |
|---|---|---|
| Critical regression (≥15pt drop) | regression_critical | critical |
New critical severity finding |
new_critical | critical |
| Deployment health check failing | deployment_down | critical |
| Regression (5-14pt drop) | regression | high |
| Sustained degradation (< 70 for 3+ scans) | sustained_degradation | high |
| Systemic issue (2+ repos same finding) | systemic | medium |
| Portfolio health drops below 75 | portfolio_warning | medium |
Digest Markdown Format
# Pi-SEO Health Digest — {date}
**Portfolio Health: {score}/100** ({delta:+d} vs previous)
## Alerts ({count})
- 🔴 CRITICAL: {project_id} — {alert_description}
- 🟡 HIGH: {project_id} — {alert_description}
## Project Scores
| Project | Score | Delta | Trend |
|---------|-------|-------|-------|
| pi-dev-ops | 87 | +2 | ↑ stable |
| ... | ... | ... | ... |
## Systemic Issues ({count})
- `{scan_type}:{title}` — affects {n} repos: {repo_list}
## Recommended Actions
1. {action}
Output Format
{
"digest_timestamp": "ISO-8601",
"portfolio_health": 82,
"portfolio_delta": -3,
"project_scores": {
"pi-dev-ops": {
"overall": 87,
"by_type": {
"security": 90,
"code_quality": 85,
"dependencies": 88,
"deployment_health": 85
},
"delta": 2,
"trend": "stable"
}
},
"regressions": [
{
"project_id": "string",
"scan_type": "string",
"previous_score": 85,
"current_score": 68,
"delta": -17,
"severity": "critical"
}
],
"systemic_issues": [
{
"key": "security:eval() usage",
"scan_type": "security",
"title": "eval() usage",
"affected_projects": ["project-a", "project-b"],
"count": 2
}
],
"alerts": [
{
"type": "regression_critical|new_critical|deployment_down|regression|sustained_degradation|systemic|portfolio_warning",
"project_id": "string",
"message": "string",
"severity": "critical|high|medium"
}
],
"digest_markdown": "string"
}