lighthouse-monitor 0.1.0

Lighthouse Keeper — PLATO fleet monitoring and beacon management

lighthouse-monitor

Automated monitoring and alerting for the Cocapn fleet. Watches services, detects anomalies, and signals alerts across the fleet — like a lighthouse beacon for distributed systems.

Architecture

lighthouse-monitor/
├── src/              ← Monitoring logic
├── config/           ← Alert rules and escalation policies
└── README.md         ← You are here

Features

• Service Health Monitoring: Periodic checks on fleet services (keeper, agent-api, holodeck, seed-mcp)
• Anomaly Detection: Configurable thresholds for response time, error rates, and availability
• Beacon System: Fleet-wide alert propagation — when one lighthouse sees trouble, all lighthouses know
• Escalation Policies: Tiered alerts (warn → critical → emergency) with configurable routing
• Integration: Works with brothers-keeper for automated remediation

Fleet Role

lighthouse-monitor pairs with:

• brothers-keeper — Automated remediation (restart, scale, heal)
• lighthouse-keeper — Agent learning and challenge suite
• fleet-mechanic — Diagnostics and repair scripts

Together they form the fleet's immune system: monitor (lighthouse-monitor) → diagnose (fleet-mechanic) → heal (brothers-keeper).

License

MIT