Reproducibility bundle (v1.9p, no‑figs) for "Scattering‑only synthesis of the proton charge radius with a guard‑banded meta‑interval"

Scope. Public reproducibility bundle (v1.9p, no‑figs) for the PRC submission. It fixes every policy number and table used in the manuscript and reproduces the seed‑and‑guard engineering interval for the proton charge radius rEr_ErE from a single, deterministic pipeline. Policy summary (fixed numbers).Seed hull H=[0.831, 0.847]H=[0.831,\,0.847]H=[0.831,0.847] fm; quantile q=0.90q=0.90q=0.90; floor g0=0.00300g_0=0.00300g0=0.00300 fm; split‑sweep quantiles Q0.90A=0.00270Q^A_{0.90}=0.00270Q0.90A=0.00270 fm and Q0.90B=0.00303Q^B_{0.90}=0.00303Q0.90B=0.00303 fm; hence g\*(0.90)=0.00303g^\*(0.90)=0.00303g\*(0.90)=0.00303 fm and the final unrounded interval [0.82797, 0.85003][0.82797,\,0.85003][0.82797,0.85003] fm (reported as rE∈[0.828, 0.850]r_E\in[0.828,\,0.850]rE∈[0.828,0.850] fm).Interpretation. The reported interval is an engineering summary interval (policy‑labeled envelope), not a frequentist CI or Bayesian credible interval; the coverage‑style table is a methodology diagnostic only. What’s inside (single source of truth). policy/policy_values.json — canonical constants used in the paper (seed hull, QqAQ^A_qQqA, QqBQ^B_qQqB, g\*g^\*g\*, unrounded endpoints, rounding policy). policy/quantiles_split_sweep.{csv,json} — split‑sweep quantiles and derived intervals (mirrors the manuscript’s policy table). results/reported_table_III.csv — transcription of the coverage‑style diagnostic reported in the manuscript’s coverage table (legacy filename retained; public scripts read this table and do not re‑run the bootstrap). data/1.1GeV_table.txt, data/2.2GeV_table.txt — PRad reduced‑cross‑section tables used by the runner. scripts/run_reproducer.py — verification, policy dump, LaTeX macro emission;env/environment.yml — pinned environment;results/joint_sweep_summary.json / results/joint_sweep_samples.csv — joint A×BA\times BA×B checks showing the max‑rule envelope bounds the joint shifts (no overshoot events). manifest.sha256.json — SHA‑256 digests for auditability. How to reproduce. conda env create -f env/environment.yml && conda activate prc-repro python scripts/run_reproducer.py --verify python scripts/run_reproducer.py --dump-policy # optional: python scripts/run_reproducer.py --emit-tex-macros python scripts/run_reproducer.py --figs-kinematics Versioning and “no‑figs.” This no‑figs bundle omits pre‑rendered figure PDFs; the runner deterministically regenerates kinematics‑only plots into artifacts/figures/. A few helper files may retain legacy version strings from earlier snapshots (e.g., v1.6p/v1.8p); policy numbers are identical and are authoritatively fixed by policy/policy_values.json. Licensing. Code: MIT. Data: CC BY 4.0. Please cite both the manuscript and this dataset (record DOI). Keywords. proton charge radius; elastic epepep scattering; low‑Q2Q^2Q2; guard‑band; engineering interval; reproducibility; PRad; radiative corrections; two‑photon exchange.