feat(worker): add generic Prometheus metrics to WorkerRunner + wire workers

[extreme4all]

Summary

Adds generic Prometheus metrics to the shared WorkerRunner framework and wires a metrics server into the framework-based workers.

Changes

Framework (components/bot_detector/worker/)

• New metrics.py: 7 generic metrics + start_metrics_server(port) (idempotent, no-op under ENVIRONMENT=test)
• core.py: WorkerRunner gains a required worker_name param; _consume() emits the metrics
• Exported start_metrics_server from the component

Generic metrics (all labeled by worker):

Metric	Type	What
worker_messages_consumed_total	Counter	messages pulled from queue
worker_messages_committed_total	Counter	messages committed (processed)
worker_messages_requeued_total	Counter	requeued msgs by reason (worker_error/requeue/error/cancelled)
worker_batch_size	Histogram	batch size distribution
worker_handle_duration_seconds	Histogram	time in handle()
worker_batch_errors_total	Counter	exceptions in handle() by error_type
worker_poll_idle_total	Counter	empty polls

Workers (each: start_metrics_server() at startup + worker_name)**

• worker_hiscore → "hiscore"
• worker_report → "report"
• worker_ban_migration → "ban_migration" (also refactored its base metrics.py to drop the duplicate start_http_server, now centralized)

Projects

• Added prometheus-client==0.22.0 to worker_hiscore and worker_report (+ re-locked). worker_ml skipped (doesn't use the worker component); worker_ban_migration already had it.

Validation

• ruff check / ruff format --check: clean
• pytest: 127 passed
• Lockfiles: only functional change is +prometheus-client (rest is uv lock reformatting)

Related

• k8s scraping config: Add prometheus metrics (Service + ServiceMonitor) for highscore-worker, worker-report, worker-ml, scrape-task-producer bot-detector-k8s#938