refactor: migrate worker_report to Worker/WorkerRunner pattern

[extreme4all-ai]

Summary

• Migrate worker_report from manual consume-loop + error-queue pattern to the reusable Worker/WorkerRunner abstraction, following the pattern established in worker_hiscore
• Extract config into pydantic-settings (settings.py), add single-record adapter (adapter.py), and encapsulate batch insert + worker logic (worker.py)
• Rewrite main.py to wire ReportWorker to WorkerRunner with graceful shutdown via stop_event and configurable N_WORKERS
• Add worker component brick to projects/worker_report/pyproject.toml
• Replace placeholder test with proper tests for adapter, insert_batch, and handle