gateway-metrics-parser

A lightweight CLI tool to parse, analyze, and visualize OpenClaw gateway performance metrics. Reads from CSV logs (generated by gateway-metrics.sh) and provides quick insights into gateway health, response times, resource usage, and uptime trends.

Installation

npm install -g gateway-metrics-parser

Usage

# Parse metrics from file
gateway-metrics parse /path/to/metrics.csv

# Show summary statistics
gateway-metrics summary /path/to/metrics.csv

# Find anomalies (high response times, RSS spikes)
gateway-metrics anomalies /path/to/metrics.csv

# Export to JSON for integration with other tools
gateway-metrics json /path/to/metrics.csv

# Show p95/p99 response times
gateway-metrics percentiles /path/to/metrics.csv

# Monitor for restarts (connection resets)
gateway-metrics restarts /path/to/metrics.csv

# Real-time tail (last N records)
gateway-metrics tail /path/to/metrics.csv --lines=20

# Help
gateway-metrics --help

Examples

Parse and summarize metrics

gateway-metrics summary metrics.csv

Output:

Gateway Metrics Summary
=======================
Records analyzed: 1,234
Time range: 2026-02-24 10:00 to 2026-02-24 18:30
Uptime: 99.7%

Response Time (ms):
  Min: 12ms
  Max: 487ms
  Avg: 45.3ms
  P95: 120ms
  P99: 310ms

Memory (RSS MB):
  Min: 48MB
  Max: 156MB
  Avg: 92.1MB
  Peak spike: +24MB at 14:15

Active Connections:
  Min: 2
  Max: 18
  Avg: 8.3

Restarts detected: 0

Find anomalies

gateway-metrics anomalies metrics.csv --threshold=200

Identifies response time spikes, memory leaks, and connection resets.

Export to JSON for dashboarding

gateway-metrics json metrics.csv > metrics.json

Useful for feeding into monitoring dashboards or Grafana.

CSV Format

Expected input (from gateway-metrics.sh):

timestamp,response_time_ms,rss_mb,connections,uptime_seconds
2026-02-24T10:00:01Z,34,92,5,36000
2026-02-24T10:00:31Z,42,93,6,36030
...

Features

• ✅ Lightweight (zero dependencies)
• ✅ Fast parsing (handles 10k+ records instantly)
• ✅ Statistical analysis (min, max, avg, percentiles)
• ✅ Anomaly detection (response time spikes, memory trends)
• ✅ Uptime calculation with restart detection
• ✅ JSON export for integration
• ✅ Human-readable summary reports

Use Cases

• Monitor OpenClaw gateway health during CI/CD pipelines
• Detect performance degradation early
• Correlate gateway metrics with application errors
• Build SLO dashboards
• Audit resource usage trends

License

MIT