Investigate heatmap view for controls

Summary

Current risk assessment uses broken hardcoded thresholds:

• hardcoded values
• Warnings over-penalize small datasets
• No context for namespace size

Solution

Replace with volume-aware algorithm:

Risk Logic

• Small (≤10 checks): 1-2 fails = Low, 4+ fails = High
• Medium (11-50 checks): ≤2 fails + ≤15% rate = Low, ≤7 fails + ≤35% rate = Medium
• Large (50+ checks): Use statistical percentiles from dataset

Weight Formula

Failures × 5 + Warnings × 0.5 (warnings much less impactful)

Acceptance Criteria

• calculates risk score relative to volume size
• Use only latest scan data per source
• Handle empty datasets gracefully

example

Input	Current	Expected
2 pass, 1 fail	High	Low
10 pass, 20 warn, 0 fail	Medium	Safe
50 pass, 25 fail	High	High