Data and code for: "Separating scalar and Ricci curvature in causal-set 𝑘 k-chain estimators: an empirical test across four geometries"

Reproducibility package for “Separating scalar and Ricci curvature in causal‑set k‑chain estimators: an empirical test across four geometries” (Glenn Pellegrin). The archive contains two fully aligned components. (1) The curved‑programme materials: Poisson‑sprinkling scripts and raw outputs for four conformally‑flat geometries (de Sitter, anti‑de Sitter, radiation‑dominated FLRW, dust‑dominated FLRW), together with the scoring and fitting code and the preregistered prediction bands used in the paper. (2) The companion flat‑space five‑arm proper‑time estimator comparison, whose measured noise‑theory inputs (a₂, b₂, a₃, b₃, ρ) are reused directly in the curved analysis and reproduce the finite‑density bias cb=−52.800±1.890 (χ² = 3.40/3) across all four densities. All results reported in the manuscript were independently regenerated from this exact data and code prior to deposit. Code is released under the MIT licence; documents under CC‑BY‑4.0. Each archive includes a README detailing file‑level provenance and step‑by‑step reproduction instructions.