{"classifications":[{"opaque_id":"SUB020","substrate":"COMPUTATIONAL_INFORMATION","reason":"Formal program representations, computability analysis, and evidence-state routing define the intervention."},{"opaque_id":"SUB105","substrate":"GOVERNANCE_PROCESS","reason":"Auditors’ precommitments, review steps, and bounded update rules perform the causal work."},{"opaque_id":"SUB109","substrate":"GOVERNANCE_PROCESS","reason":"A structured mentor relationship, debriefs, and autonomy rules transmit the intended judgment."},{"opaque_id":"SUB122","substrate":"COMPUTATIONAL_INFORMATION","reason":"Versioned mapping packages, context routing, namespaces, and state-transition logic are primary."},{"opaque_id":"SUB048","substrate":"COMPUTATIONAL_INFORMATION","reason":"An explicit configuration tree and rule-based branch-filtering procedure allocate replay computation."},{"opaque_id":"SUB033","substrate":"GOVERNANCE_PROCESS","reason":"Defined authority, permitted actions, feedback review, and escalation govern exception resolution."},{"opaque_id":"SUB042","substrate":"COMPUTATIONAL_INFORMATION","reason":"A data-linking and network-motif detection layer produces the reviewable signals."},{"opaque_id":"SUB034","substrate":"GOVERNANCE_PROCESS","reason":"A coordinator-managed intake, release, expiry, and escalation workflow regulates reviewer work."},{"opaque_id":"SUB119","substrate":"MEASUREMENT_INSTRUMENTATION","reason":"Multi-sensor spatial-temporal sensing and corroborating measurement identify the forming condition."},{"opaque_id":"SUB089","substrate":"GOVERNANCE_PROCESS","reason":"Guardrails, dominance review, preference authority, and recorded selection rules govern the choice."},{"opaque_id":"SUB092","substrate":"COMPUTATIONAL_INFORMATION","reason":"A representation-independent behavioral contract and black-box conformance logic determine substitution."},{"opaque_id":"SUB134","substrate":"GOVERNANCE_PROCESS","reason":"Reverse planning, dependency mapping, commitments, and accountable review create the intervention."},{"opaque_id":"SUB084","substrate":"COMPUTATIONAL_INFORMATION","reason":"Request normalization, exact-match coalescing, token admission, and scheduling logic regulate demand."},{"opaque_id":"SUB026","substrate":"GOVERNANCE_PROCESS","reason":"A time-bounded control-repair workflow with gates, authority limits, and stop rules is primary."},{"opaque_id":"SUB040","substrate":"GOVERNANCE_PROCESS","reason":"Published service commitments, pricing, and self-selection rules allocate facility capacity."},{"opaque_id":"SUB117","substrate":"COMPUTATIONAL_INFORMATION","reason":"A constraint-filtered, capacity-aware routing algorithm selects review destinations."},{"opaque_id":"SUB021","substrate":"COMPUTATIONAL_INFORMATION","reason":"Frozen predictions, residual computation, attribution records, and model-update logic drive learning."},{"opaque_id":"SUB045","substrate":"GOVERNANCE_PROCESS","reason":"Authorized weighting, eligibility floors, sensitivity review, and approvals govern the decision."},{"opaque_id":"SUB103","substrate":"COMPUTATIONAL_INFORMATION","reason":"A versioned predictive model, residual comparator, and threshold gate drive reviewer triage."},{"opaque_id":"SUB115","substrate":"COMPUTATIONAL_INFORMATION","reason":"A formal material-lineage representation and partition-based inventory calculation enforce valid aggregation."}]}