quantum-governance-testbed: Preregistration v1 — QG-001, QG-002, QG-004 (locked before hardware access)

Preregistered experiments testing whether governance verdicts (PASS/HOLD/FAIL) produced by imperfect quantum hardware are trustworthy enough to cross an execution boundary. Part of the Remnant Fieldworks / Coherent Inheritance Framework / ExecutionProof program. This is a preregistration-only deposit: questions, thresholds, pass rules, and kill conditions are SHA-locked BEFORE any IBM Quantum hardware result is computed. Results will be published in subsequent versions regardless of verdict. QG-001: Verdict Stability Under Hardware Noise — same circuit, same backend, five noise treatment conditions (raw, dynamical decoupling, measurement mitigation, gate twirling, ZNE). QG-002: Cross-Backend Verdict Reproducibility — same circuit on 2–3 independent IBM processors. QG-004: Mitigation Cannot Manufacture Truth — can error mitigation flip a verdict against a deliberately falsified analytic target? Substrate: DM-002 sterile neutrino oscillation circuit (single-qubit, known analytic answer, clean PASS on ibm_marrakesh). Positive control: BELLWETHER and TRINITY series (CHSH Bell-violation witnesses). GitHub: derekhone/quantum-governance-testbed