Gravitational Manipulation Theory v3.1 — H5 Post-Approval E1–E7 Execution Attempt and Fail-Closed Audit Record

This dataset records the single post-approval H5 reveal and E1–E7 execution attempt for the Gravitational Manipulation Theory (GMT) v3.1 synthetic mechanism-verification program. The execution was derived from the externally fixed H5 blinded prospective pre-registration: https://doi.org/10.5281/zenodo.22153766 Before execution, the published H5 specification, encrypted package, Custodian Freeze Manifest, and Integrated Harness Freeze Bundle were downloaded from Zenodo and verified against their registered SHA-256 values. The H5 specification commitment was: 434887c06db2dfe2f464c101d62424738f785fa351afacaf6079d07d24d83727 The frozen Integrated Harness Bundle SHA-256 was: 12f3c7d2dd1168df38f5b641febed93d0b62391d6e4bafb69ed26bc94b0b2732 The single post-approval reveal completed with status REVEAL_VERIFIED. The encrypted package, authenticated plaintext package, seed-root commitment, 54-cell manifest, and six designated null controls were verified against the pre-registered commitments. The complete manifest was deterministically re-derived from the authenticated seed root before Rev.G execution was permitted. The registered E1 execution processed all 54 expected cells. Chain continuity, checkpoint integrity, manifest identity and ordering, and the information-firewall checks passed. However, the pre-registered NULL_CONTROLS_OK validity gate failed because at least one designated null control qualified outside the frozen null-control condition. Under the locked validity-first decision semantics, the E1 result was: mechanism_verdict = PASSscientific_run_status = INVALID_RUNE1 end-to-end verdict = FAILterminal_state = INVALID_RUNtyped_error = ValidityErrorreason = null control qualifiedpredicate_evaluations = 0official_decision_evaluations = 0terminal run seal = not issued INVALID_RUN is not TRANSFER_FAILURE and is not a scientific result. No TRANSFER_SUCCESS, TRANSFER_FAILURE, or NULL_CONSISTENT scientific terminal state was calculated for H5 E1. The remaining registered mechanism tests produced: E2 — R0 reproduction-only commitments: PASSE3 — sealed-record, chain-link, and ordering tamper detection: PASSE4 — preflight fault matrix: PASSE5 — mid-run fail-closed fault handling: PASSE6 — frozen three-state decision vectors: PASSE7 local artifact audit: PASSE7 cross-environment R1: PENDING_EXTERNAL_REPLICATION The overall H5 harness status was FAIL_OR_INCOMPLETE. The canonical H5 result-record digest was: 673cfb4ee99e180a98af3264845541eb3e50b82e2d4b240ae157890ecf82507c The public result JSON file SHA-256 was: 251f10f878e715aee844846784712540c88290c0d2af326b3c19da5d94f8fe8f The deposited ZIP SHA-256 is: 1b22b72543c940ce9f72e482eb44c2a33e7682c059ef44e6a79f832aa4bf5c6f Both the reveal-attempt marker and the E1-attempt marker were created. H5 is therefore closed as a no-retry execution attempt. No threshold, rule, manifest, policy, or code version was changed in response to the observed null-control outcome. The deposited public dataset excludes the custodian key, plaintext seed root, individual seed leaves, seed paths, and plaintext manifest. This record concerns a synthetic representative mechanism-verification workflow using TEST_ANCHOR and test approver keys. It does not establish physics-kernel integration, experimental confirmation, measured-data validation, independent certification, security certification, production qualification, cross-environment R1 completion, or completion of GMT v3.1.