MSC-5 Pre-Registration Manifest · OHV-Tomography 2-Chip Falsification Test on IBM Heron r2 (kingston + marrakesh)

Pre-registered submission package for the MSC-5 Orbit Hamming Variance (OHV) tomography cross-chip experiment on IBM Heron r2 quantum hardware (ibm_kingston + ibm_marrakesh, 2-chip scope after open-instance plan amendment dropped ibm_torino). 23 SHA-256-locked files: 9 production scripts (dynamic backend.properties() bulk-spine selector; Latin-square shot-interleaved submitter; post-transpilation OHV verifier; density-matrix Lindblad null simulator; reservoir-correction Kraus construction with machine-precision F-fix verification) + 14 pre-registration documents (main MSC-5 pre-reg with single primary test + falsification commitment; parallel MSC-5b controlled-noise track; theory: strict-sensitivity inequality between OHV and standard cycle-benchmarking observables on engineered channel class; explicit Kraus construction appendix; relation-to-prior-work positioning vs Shor-Laflamme weight enumerators / Krawtchouk transforms / XEB / Cycle Benchmarking / Direct Fidelity Estimation; Skeptic patches × 2; Rounds 5-6-7 specialist syntheses with 5 retractions documented). This manifest constitutes third-party-verifiable proof that all predictions and methodology were locked before any chip data was collected. The 5th + 6th pre-registrations in the MSC series; predecessors at Zenodo concept 10.5281/zenodo.19683826. Master seed 0x6909, transpiler seed 42, scoring weights (alpha=10ms, beta=10ms, gamma=2.0) all locked. Total budget 262K shots (MSC-5) + 2.49M shots (MSC-5b). Predicted pooled Delta-gap (MOD - RAND) under H_A vs H_B: CIs disjoint, decisive in single submission. Predicted MSC-5b theta_crit = pi/12 rad. Sixth retraction triggers if results contradict both H_A and H_B. Companion: KAAL-DISCIPLINE.md Truth #14 (Master Meta-Theorem (R, g, k)).