feat(analytics): implement resource-to-value correlation and hybrid attribution

[victoriacheng15]

Summary

Completes the core architectural loop of the Analytics Engine by implementing "Resource-to-Value" correlation. This update enables hybrid energy attribution across K3s and host services, enforces a sustainable data lifecycle via retention policies, and ingests business production counters (Syncs/Tasks) to enable high-fidelity Unit Economics analysis.

List of Changes

• Value Chain Integration: Implemented "Business Value Ingestion" by bridging production counters (thought/document syncs and webhook/synthetic requests) from Ingestion and Proxy services into the analytical dataset.
• Hybrid Visibility Architecture: Eliminated energy "dark matter" by implementing a CPU-share attribution model that account for non-containerized systemd services alongside Kubernetes pods.
• Data Lifecycle Governance: Enforced automated storage sustainability via a 30-day retention policy in TimescaleDB hypertables, ensuring long-term performance and manageable growth.
• Relational Integrity: Unified the mapping layer to synchronize cluster-realistic identifiers with business feature IDs, providing a consistent anchor for relational joins between costs and value.

Verification

• make test passes for internal/analytics verifying hybrid attribution and value unit recording.
• Database schema verified via EnsureSchema to include retention policies and metric_kind enums.
• Manual verification of analytics_metrics table confirms recording of energy, cost, carbon, and value_unit kinds.