v1.2.0

Added

• --verbose / -v CLI flag to enable DEBUG-level console logging (default remains INFO). Surfaces per-workload skip/ignore decisions and other diagnostic detail.
• Aggregate INFO Discovery summary: <N> included, <N> skipped (--skip-containers workload-name match), <N> container-skipped (--skip-containers container-name match), <N> unparseable images, <N> not in manifest, <N> inspected log line emitted after each cluster discovery cycle. The five buckets are mutually exclusive and sum to inspected.
• Per-workload DEBUG log lines explaining why each workload was dropped from the verification scope: container-name skip, unparseable image, anchor-not-found in image path, or component-not-in-manifest. Deduplicated per (workload, image).

Changed

• Workload discovery performance: ReplicaSets and Pods are now fetched once per discovery cycle using paginated limit / _continue requests and served from in-memory maps. For a large namespace (e.g. 92 Deployments / 848 ReplicaSets), discovery drops from ~390+ scattered Kubernetes API calls per cycle to roughly 3 paginated namespace-wide reads.
• The workload object loaded during discovery is now carried forward into pod listing and the stability audit, removing redundant read_namespaced_{deployment,stateful_set,daemon_set} round trips per workload.
• Per-workload Skipping inspection of workload <name> (matched skip pattern) log demoted from INFO to DEBUG. Use the new Discovery summary line for an at-a-glance INFO view.
• Console log format: logger name field is now the literal kubernify instead of the full module path (e.g. kubernify.cli).
• Internal: construct_component_map() now returns a ComponentMapResult dataclass exposing the component map plus three exclusion-reason workload sets. No public CLI or report-schema impact.
• Internal: list_pods_by_deployment / list_pods_by_stateful_set / list_pods_by_daemon_set collapsed into a single list_pods_for_workload.

Fixed

• Eliminates HTTP 429 Too Many Requests and urllib3 IncompleteRead errors that could surface against large namespaces during cluster discovery.