COMPLETE / INDEPENDENTLY VERIFIED / ACCEPTED SPECIFICATION ONLY / NATIVE FEASIBILITY UNRESOLVED
Owner-authorized DC-M4-PARAMETER-MECHANICAL-WORK-PORT-FEASIBILITY-SPEC-001, exact entry a007e7e08acc768f08b83d96139a7cea25f0e04e. DC-M4 — Conversion-Work Ownership Contract remains an unchanged conditional contract. This task defines and verifies the prospective procedure; it does not run calibration, implement coupling or qualify biological correction.
Four independent dimensionless parameter groups are identified. Force calibration uses 80 sealed design rows: four training rows define a joint positive parameter region and 76 remain held out. The kinetic stage uses an independently observed unclipped net rate, both load signs, fixed held-out amplitudes and time-refinement evidence. A fixed selector defines probe geometry while retaining the full parameter uncertainty. Work/fuel calibration requires an independently owned force-displacement donor and A/W observations. Synthetic observations never identify production parameters; all four remain unassigned.
S0–S5 bind provenance, independent anchors, force/work ownership, native API compatibility, solver/contact/funding/staging obligations and explicit stop rules. Certified uncertainty must meet independently sealed accuracy and tightness budgets; broad intervals containing zero are insufficient. Future evidence collection and implementation require separate authorization.
Calibration anchors are missing. The naive work-port adapter is rejected: simultaneous legacy forces can store part of the coupling energy, so coupling release cannot all be classified as bath heat. Extra A loading requires an identified physical work-producing channel. Existing tension APIs reject negative tension; the capped signed external-force hook and explicit integrator do not establish exact discrete work. Native A also affects pressure and actuator funding, requiring explicit causal staging. This rules out direct adapters without proving every possible realization impossible.
Exact rational work-balance and reduced two-spring counterexamples are diagnostic mathematical witnesses. They do not verify a native MaterialMesh solver, calibration source, runtime rollback, contact accounting or checkpoint lifecycle. The floating net-rate inversion check is synthetic smoke only. R13 damping remains uncorrected; R4 production reference and R10 diagnostic-only; no automatic successor.
Isolated checkout /home/sketch/Projects/digital_cell-work-port-feasibility-spec; local evidence /home/sketch/Projects/digital_cell-work-port-specification-evidence/; additive Atlas root /srv/ATLAS/100_ACTIVE/Projects/DIGITAL_CELL/evidence/work-port-specification/. Existing work and PR44–47 remain preserved. Final exact-head CI/artifact, preservation and Notion readback receipts follow publication.
Exact head 101e635d6498b9cfe451b2d702d64a4440939963; draft PR48; CI39075467602 PASS. Downloaded artifact 12116892436, ZIP sha256:951fdcc08d3ce1d49cf82d1201d0a166582f67df98a33804459256bebd6fa155. All 38/38 repository manifest bindings and 22/22 frozen inputs verify against the exact head, including 16 scientific source files and the unchanged accepted contract package. Hosted verification reproduces the local exact-head result byte for byte; manifests match semantically.
Independent five-file hash-bound review is ACCEPTED_SPECIFICATION_ONLY. Parent checks pass 30 exact mathematical simultaneous-force cases and 14 reference control groups. The prospective force design has 80 rows with four fixed joint-training and 76 held-out roles; these observations have not been collected. The floating net-rate inverse remains synthetic smoke only. Future contact, native solver, staging, lifecycle and rollback requirements have not been executed. Historical governance failures remain identical to the entry baseline; scoped new records pass.
Seven prior local checkouts and PR44–47 retain their original heads, bases and existing work. Production parameters remain unassigned, naive direct adapters are rejected, and whole-candidate feasibility remains unresolved. Next prerequisite: an independently owned mechanical source/receiver architecture and independent calibration evidence. No automatic successor.
Receipts are retained outside Git at /home/sketch/Projects/digital_cell-work-port-specification-evidence/ and additively under /srv/ATLAS/100_ACTIVE/Projects/DIGITAL_CELL/evidence/work-port-specification/ to avoid self-referential commit hashes.
Direct Atlas verification confirms the expected mounted UUID and bind, all 143 delivered publication files and their hashes, hosted ZIP integrity, and 38/38 internal repository bindings. All 373 prior scoped archive files remain hash-identical: 206 reaction-feasibility and 167 conversion-contract files. Original Atlas source, seven prior local checkouts and PR44–47 remain preserved. Receipt /srv/ATLAS/100_ACTIVE/Projects/DIGITAL_CELL/evidence/work-port-specification/archive-verification-101e635.json; local original evidence remains retained. No prior archive file was replaced or deleted.