Add domain tests and pipeline CI

[derekk024]

What changed

• Moved event validation/storage, aggregation, and alert evaluation into shared production headers so the same logic is directly testable.
• Added CTest coverage for JSON validation, SQLite event-ID idempotency, weighted packet-drop aggregation, percentile interpolation, link-quality averaging, threshold boundaries, and aggregator failures.
• Added a real-process integration test that launches ingest and aggregator against a temporary SQLite database and checks validation errors, duplicate suppression, satellite/window filtering, aggregate results, and Prometheus counters.
• Added a GitHub Actions workflow that configures, builds, and runs the complete suite on pull requests and pushes to main.
• Documented the test workflow and updated the project structure and limitations.

Why

The repository previously had no automated regression coverage, and its key rules lived inside executable entry points. This keeps the existing three-service architecture while making those rules independently testable and adding an end-to-end check across the real HTTP and persistence path.

Validation

• Debug build and all 4 CTest entries passed locally.
• Warning-enabled Release build (-Wall -Wextra -Wpedantic) and all 4 CTest entries passed locally.
• The service pipeline integration test passed 10 consecutive runs.
• python3 -m py_compile tests/service_pipeline_test.py
• git diff --check