This record provides the reproducibility package accompanying the manuscript: “A Fixed-Shape Gravitational Kernel for Galaxy Rotation Curves: An Empirical Tool without Per-Galaxy Tuning.” The archive contains a fully executable Python codebase that reproduces all galaxy-scale numerical results presented in the paper using the publicly available SPARC rotation-curve database (175 disk galaxies). What the code does Loads SPARC rotation-curve decompositions (*_rotmod.dat) Implements a fixed-shape gravitational response kernel Fits kernel amplitudes in oracle mode (using observed kinematics) Calibrates a global predictive amplitude law based on a coherence parameter Reproduces: full-sample RMS statistics (oracle and predictive modes) representative rotation-curve figures amplitude–coherence scaling relations The package also includes optional scripts to reproduce the cluster-scale oracle checks presented in the paper, given appropriate cluster profile inputs. Scope and intent This code is designed as a mechanical empirical tool, not a simulation framework or cosmological model. It makes no claim to predict galaxy dynamics from baryonic matter alone; instead, it tests whether a universal response kernel combined with a global coherence-dependent amplitude can reproduce observed rotation-curve structure without per-galaxy tuning. Reproducibility The code has been executed end-to-end against the full SPARC database. All figures and tables in the manuscript are reproducible with this package. No internet access is required once the SPARC data are provided locally. Contents Source code implementing the kernel and fitting procedures Scripts to reproduce all SPARC figures and summary tables Scripts for optional cluster oracle tests Exact paper figures and tables included for reference Requirements Python ≥ 3.9 with standard scientific libraries (numpy, pandas, matplotlib). The SPARC database must be obtained separately from its official public source.