| name | capacity-monitor |
| description | 容量监控 — 持续监控磁盘、内存、进程数等系统容量指标,提前预警 |
| version | 1.0.0 |
| license | MIT |
| metadata | {"author":"blackplume233","actant-tags":"capacity,monitoring,alerting,scavenger","layer":"auxiliary","agent":"scavenger"} |
Capacity Monitor
Overview
Continuous monitoring of disk, memory, process count, and other system capacity metrics; early warning.
Threshold Configuration
- Define per-metric thresholds (e.g., disk > 85%, memory > 90%)
- Support absolute values (e.g., free disk < 5GB) and percentages
- Configurable per workspace or global defaults
Trend Prediction
- Track metric history; extrapolate growth rate
- Predict when threshold will be breached (e.g., "disk full in ~7 days")
- Use simple linear regression or moving average
Alert Escalation Levels
| Level | Condition | Action |
|---|
| Info | Approaching threshold (e.g., 70%) | Log only; optional notification |
| Warning | At threshold (e.g., 85%) | Notify user; suggest cleanup |
| Critical | Severe (e.g., 95%) | Urgent alert; integrate with Maintainer for automated response |
Integration with Maintainer
- On critical: delegate to Maintainer for automated cleanup (resource-scan + cleanup-policy)
- Pass context: which metric, current value, suggested actions
- Maintainer reports back; Capacity Monitor logs resolution