Governance-First Predictive Modelling for 'Gazepoint' Research R package gp3ml version 0.1.0 Provides governance-first infrastructure for leakage-resistant predictive modelling and validation using 'Gazepoint'-derived research data. Supports explicit task and role declarations, feature-provenance manifests, group-aware holdout splitting and repeated resampling, fold-local preprocessing, governed model engines, discrimination and calibration metrics, bootstrap uncertainty, external-validation reports, model cards, and reproducibility reports. Intended only for explicitly observed, non-sensitive outcomes and declared scientific purposes. Use is prohibited for person identification, biometric authentication, health or protected-attribute inference, and direct or indirect inference of emotion, stress, personality, deception, cognition, comprehension, intent, or other mental states. Software Depends: R (>= 4.1.0) Suggests: jsonlite, keras3, knitr, nnet, pkgdown, ranger, rmarkdown, testthat (>= 3.0.0), xgboost Stefanos Balaskas <s.balaskas@ac.upatras.gr> Comprehensive R Archive Network (CRAN) Stefanos Balaskas [aut, cre] (ORCID: <https://orcid.org/0000-0003-2444-9796>) MIT + file LICENSE (https://CRAN.R-project.org/package=gp3ml/LICENSE) 2026-07-30 application/tgz https://CRAN.R-project.org/package=gp3ml doi:10.32614/CRAN.package.gp3ml en-US