Supplementary evidence and reproducibility materials for "Assuring History-Dependent AI Decision Support: Cold-Start Reliability, Metric Coherence, and Queue Stability"

This record contains the supporting evidence and compact reproducibility materials for the manuscript “Assuring History-Dependent AI Decision Support: Cold-Start Reliability, Metric Coherence, and Queue Stability,” prepared for submission to Information Systems Frontiers. The repository contains supplementary methods and results (`ESM_1.pdf`), a machine-readable aggregate evidence workbook (`ESM_2.xlsx`), and an aggregate reproducibility artifact with deterministic validation/regeneration code (`ESM_3.zip`). The release supports aggregate claim tracing, released-ledger recalculation, inspection of declared design invariants, and regeneration of the principal released tables and figures. It does not redistribute raw SECOP II source bytes, linkable row-level analytical records, fitted model or calibration objects, original row-level resampling inputs, credentials, or private continuity records, and it is not represented as unrestricted raw-source-to-model reproduction. The research was conducted independently in the author's personal academic capacity; the employer had no role in study conception, source acquisition, analysis, interpretation, manuscript preparation, or the decision to submit.