feat: Add Prometheus metrics export endpoint

[Patrick-Ehimen]

Summary

Expose metrics in Prometheus format to enable integration with standard monitoring stacks (Prometheus, Grafana, Datadog).

Background

The MetricsCollector already tracks useful metrics internally. Exposing these in Prometheus format enables production observability without custom tooling.

Acceptance Criteria

• Add /metrics endpoint with Prometheus text format
• Export existing metrics from MetricsCollector:
  • lighthouse_auth_total{status="success|failure"}
  • lighthouse_cache_hits_total
  • lighthouse_cache_misses_total
  • lighthouse_cache_size
  • lighthouse_request_duration_seconds{operation="upload|download|..."}
  • lighthouse_security_events_total{type="..."}
• Add histogram for request latencies
• Include process metrics (memory, uptime)
• Document metrics in README

Example Output

# HELP lighthouse_auth_total Total authentication attempts
# TYPE lighthouse_auth_total counter
lighthouse_auth_total{status="success"} 1542
lighthouse_auth_total{status="failure"} 23

# HELP lighthouse_cache_hit_rate Cache hit rate
# TYPE lighthouse_cache_hit_rate gauge
lighthouse_cache_hit_rate 0.87

Related Files

• apps/mcp-server/src/auth/MetricsCollector.ts
• apps/mcp-server/src/server.ts