Skip to main content

Aggregation and Risk Synthesis

The aggregation layer is where module-native outputs become category scores, risk summaries, fast-fail packets, and TRI-facing report artifacts.

Artifacts involved in aggregation​

The example bundle exposes the aggregation chain through:

  • results/combined_unified_computational_outputs.json
  • thresholds/category_scores.json
  • thresholds/measurement_profile.json
  • thresholds/risk_summary.json
  • results/risk_channels_map.json
  • results/cbtri_policy_v1.json
  • results/summary.json

What aggregation does​

Aggregation combines:

  • raw module outputs
  • category policy
  • threshold logic
  • risk-channel mapping
  • claim and tier rules

into the final report-facing artifacts.

Why this is not a black box​

The example output keeps the aggregation inputs inspectable:

  • category scores show individual metric-level risk assignments
  • measurement profile shows threshold rules
  • risk summary shows category-by-category breakdown
  • risk channels map shows how modules feed report channels
  • CB-TRI policy shows channel weighting and band thresholds

That makes the aggregation layer auditable rather than purely narrative.