Add metrics and monitoring

[davesleal]

Description

Implement observability features to track bot performance, usage, and costs.

Proposed Metrics

Usage Metrics

• Messages processed per hour/day
• Investigations triggered
• PRs created
• Commands executed
• Active channels/users

Performance Metrics

• Claude API response time
• App Store Connect API latency
• Session memory usage
• Uptime and availability

Cost Metrics

• Claude API tokens used
• Estimated cost per day/month
• Cost per investigation
• Cost breakdown by channel

Business Metrics

• App Store reviews processed
• Bugs auto-triaged
• Time saved (estimated)

Implementation Options

Option 1: Built-in (Simple)

# Simple JSON logging
metrics_logger.log({
    "timestamp": now(),
    "event": "investigation_complete",
    "tokens_used": 1234,
    "cost": 0.05
})

Option 2: Prometheus + Grafana

• Expose /metrics endpoint
• Track counters, histograms, gauges
• Visualize in Grafana

Option 3: Cloud Monitoring

• AWS CloudWatch
• Google Cloud Monitoring
• Datadog

Benefits

• Cost visibility
• Performance optimization
• Usage insights
• Capacity planning