Define CalibrationControlStatus enum (pathogenic, benign)

[bencap]

Summary

Add a CalibrationControlStatus enum to models/enums/ that captures the two ACMG-coded clinical significance values required for calibration controls.

Background

Calibrations in MaveDB are derived from empirical controls — variants with independently known clinical significance. To store these controls, we need a controlled vocabulary for clinical status. The design decision is a strict 2-tier ACMG-aligned enumeration (pathogenic, benign) rather than the full 5-tier classification, reflecting the minimum information needed to anchor a calibration.

Proposed Behavior

A new CalibrationControlStatus enum with values:

• pathogenic — the variant is classified as pathogenic per ACMG criteria
• benign — the variant is classified as benign per ACMG criteria

The enum should follow the existing pattern in models/enums/ (e.g., FunctionalClassification in models/enums/functional_classification.py) and use a string-based SQLAlchemy enum with native_enum=False and validate_strings=True.

Acceptance Criteria

• CalibrationControlStatus enum is defined in src/mavedb/models/enums/calibration_control_status.py
• Enum values are pathogenic and benign (lowercase strings)
• Enum uses the same SQLAlchemy-compatible pattern as FunctionalClassification
• Enum is importable and used in the CalibrationControl model (follow-on schema issue)

Implementation Notes

• Follow the pattern in models/enums/functional_classification.py
• The enum intentionally excludes intermediate tiers (VUS, Likely Pathogenic, Likely Benign) — this is by design for the minimalist controls data model