TraceQueue: Source-Grounded Triage Accelerates Paper–Repository Verification

Between repository resolution and rerun execution, an auditor must turn one queue entry into a verified follow-up decision. TraceQueue makes this handoff measurable: $1,085$ resolved pairs share an ordered paper-plus-repository evidence frame, a $42$-field disclosure contract, calibrated risk and abstention, field-level source pointers, and a temporal rerun ledger. With GPT-4.1, evidence, temperature, calibration folds, scalar request, and thresholds held fixed, replacing a free-form rationale by the complete contract yields $\Delta$ROC $=-0.003$ ($95$ CI $[-0.018,+0.013]$), improves ECE by $0.008$ ($p=0.028$), and increases explicit abstention by $5.4$ points ($p=0.003$); 92.8% of audited Open Structured pointers identify their intended source. The decisive test is a blinded study of 126 pairs: the structured entry reduces median verification time from 17.8 to 13.6 minutes (23.6%, $p=0.004$), lowers false follow-up from 14.6% to 8.8%, and raises verified-issue recall by 20 minutes from 64.9% to 71.6%. Paper and table evidence adds $0.055$ holdout AUC over repository-only features and retains $0.048$ after maturity balancing; in a fresh $204$-pair cohort, full-contract risk has adjusted odds ratio 1.62 for decision-changing verified issues. Cross-team and disjoint reruns preserve the score ordering. These results show that ranking quality alone can conceal a consequential system property: how efficiently a prediction becomes a verified expert decision.