Post-audit Data and Code for Reappraising Flavonoid Evidence in Experimental Pulmonary Fibrosis

This version provides the corrected analytical dataset and reproducibility package for a dependency-aware reappraisal of flavonoids in experimental pulmonary fibrosis. Following source-level audit, the revised analysis includes 25 experimental study units and 175 effect records. The package contains the corrected dataset, reconciliation and exclusion logs, study-level evidence-structure summaries, meta-analysis and sensitivity outputs, risk-of-bias materials, analysis code, and editable figure sources. Although 165 of 175 records (94.3%) favoured flavonoids, the evidence was concentrated in selected compounds and outcomes. Only 10 experiments informed hydroxyproline, and its prediction interval crossed zero. The findings support a positive average signal in selected experimental settings, but not stable class-wide efficacy. The unchanged 28-report-identifier/185-record source archive is retained for provenance. The literature search remains locked through 2026-06-01.