qector-reproducibilitylisted

Reproducibility and claim boundaries for QECTOR (manual 19, 22, 27, Appendix D). Covers the chapter 22.3 required-artifact metadata, the safe / unsafe wording rule (manual 22.5), the reproduction commands in Appendix D, the competitive-harness pipeline (same Stim circuit, same samples, same DEM, same observable scoring), the withdrawn benchmark policy, and the hot/cold path label. Load for any question about how to reproduce a result, what counts as evidence, or how to write a paper-ready claim.
GuillaumeLessard/qector-claude-plugin · ★ 1 · AI & Automation · score 62

Install: claude install-skill GuillaumeLessard/qector-claude-plugin

# QECTOR Reproducibility and Claim Boundaries Source of authority: v1.0.0 reference manual, chapters 19, 22, 27; Appendix D. ## Required elements of a published result (manual 19) For every published result, the repository policy requires: - a scoped one-sentence claim - a clean git commit - an environment capture (OS, Python, Rust, package versions, hardware) - the exact command - raw JSON / CSV artifacts with SHA-256 - a Wilson interval for LER - hot / cold path labels for latency - safe / unsafe wording attached to every result ## Required artifact metadata (manual 22.3, table 22.1) | Field | Required detail | | --------------- | -------------------------------------------------------------------------------- | | code family | `repetition`, `ring`, `rotated_surface`, `toric`, `heavy_hex`, `color_code`, `bicycle`, `bivariate_bicycle`, custom (from `H_matrix`), or DEM-derived | | distance / size | `distance`, `rounds`, `n_checks`, `n_qubits`, `n_detectors`, or code parameters | | noise model | physical error rate, channel/circuit source, `code_capacity` vs `circuit_level` tag | | decoder | exact class and mode (weighted / batch / GPU flags) | | sample count | trials, shots, warmup, seed | | metric | correctness, LER, latency, throughput, memory, GPU bit-identity, or scaling | | e