Add Prometheus metrics integration

Integrate Prometheus metrics with request counts, response times, error rates, and system metrics. Add /metrics endpoint and Grafana dashboard examples.

Acceptance Criteria

• Add Prometheus metrics library to Cargo.toml dependencies
• Implement metrics collection for:
  • HTTP request count by method and status code
  • Request duration/response times (histogram)
  • Active connection count
  • Error rates by endpoint
  • System metrics (CPU, memory usage)
• Create /metrics endpoint that exposes metrics in Prometheus format
• Add middleware to automatically collect HTTP metrics
• Include example Grafana dashboard configuration
• Document metrics collection and monitoring setup
• Add Docker Compose configuration for local Prometheus/Grafana stack
• Ensure metrics collection has minimal performance impact

Technical Notes

• Use prometheus crate for metrics collection
• Consider using metrics crate as abstraction layer
• Implement proper error handling for metrics collection
• Add configuration options to enable/disable specific metric types
• Follow Prometheus naming conventions for metrics

Start by inspecting Cargo.toml and the server's HTTP entry points to understand the existing Rust request handling. Then determine where middleware and the /metrics endpoint belong, and review the Prometheus and Grafana configuration requirements. Done means the requested metrics, endpoint, dashboard example, documentation, Docker Compose stack, and configuration options are implemented with tests or verification for the acceptance criteria.