[Clear1nsight]

Publication status

This is the review PR for the completed audit contribution. Creating the corresponding PR in CrocSwap/integer-mult-bounds was attempted twice and denied by the current GitHub integration with HTTP 403, Resource not accessible by integration. This fork PR is not represented as upstream submission.

The intended upstream target is CrocSwap/integer-mult-bounds:main, with head Clear1nsight:audit/staged544-r437-exact-head-admission. The fork base review/canonical-main-20ecf289 points exactly to canonical main 20ecf2891d80b5acf18d980b2164fb44ea7faa90, avoiding unrelated upstream commits in the review diff. The scientific commit is b122e6ceab16bc1c6df818490ea5a21e864e0786. No merge or selected-frontier change is requested.

Contribution

Only research/staged544-r437-admission-audit/ is added. This is a partial admission audit, not supplier admission and not a new kappa. Frozen predecessor packages are unchanged.

The new standard-library checker imports no predecessor verifier. It independently checks PR3647's complete OQ4 template transition from 36ff441479ff9c7e70d83f0becccd44da50effb8, covered by Sagar Thakrar's PR3700, to current head a24f9ff95c12fd0b6792e3f710569237efd72db2. Both templates are retained byte-for-byte with SHA256 and Git-blob pins.

A backwards last-touch scan rederives idle windows. Explicit coordinate occupancy checks all 3,933 stage/helper slots per replica, chronology, and contiguous residual suffixes. Both complete owner maps are reconstructed, including all 125,856 helper occurrences per head and every source/target owner. Exhaustive template comparison identifies exactly 69 changed semantic leaves. The slot move changes 32 helper owner descriptors and gives residual profile delta {26:32,34:-64,42:32}, preserving child count and rank mass. Every other field, including local programs, schedule, cover, and stock declarations, is unchanged.

Seven semantic mutants are rejected. The modified-program and modified-cover controls pass the geometry gates and fail the exhaustive transition gate, guarding against silently carrying older identity coverage to a changed program. The README proves the improvement direction for every 0<c<1 by strict concavity under the unchanged inherited invoice. It does not claim discovery of the slot move or a new physical hosting theorem.

Exact-head execution evidence

The current R437 author's self-contained cold verifier passed all 13 commands in hosted Python 3.12.15 and local Python 3.13.5 executions. The current c = 1736646042038446578551/10^24 and next-grid rejection were cross-checked with the credited, unmodified PR3700 rational engine; the old profile at the new c was rejected. These are existing-code executions, separate from the new independent transition implementation.

For staged544 T385, PR3802 at de2cd5c58009205210e3bf328b0ef5cadf3903af, the published full wrapper was run in a fresh work directory, with one job, pinned dependencies, and explicitly declared authenticated native/ancestry reuse. The imposed 780-second limit was reached during discovery/audit, with exit 124. Owners, parent-lift, rebind, discovery bind, and discovery emit completed with exit 0. The retained status was PREPARED_REPLICA_SAME_WORD_DERIVATIVE_NOT_ADMITTED. No completed fresh discovery, outer stages, or release is claimed. The timeout is not a counterexample.

A subsequent fresh Python 3.12.15 run of T385's bounded verify.py passed the exact aggregate, all 1,413,720 literal helper roles, three owner corruption controls, regenerated outer ledgers, and conditional-gain checks. This authenticates stored complete-admission evidence and does not turn the timed-out full replay into a pass.

Verification

The exact scientific commit passed hosted metadata validation, registered replay, optimised-Python receipt equality, and source-history/publication-byte checks. The downloaded publication bytes also passed Git-free replay and CLI controls for altered input bytes, altered result bytes, missing inputs, and symlinked sources.

python3.12 -B scripts/research_package.py --package research/staged544-r437-admission-audit
python3.12 -O -B research/staged544-r437-admission-audit/verify.py
python3.12 -B scripts/check_research_commit.py --package research/staged544-r437-admission-audit --check-source-history

Publication and bounded T385 jobs both passed at https://github.com/Clear1nsight/integer-mult-bounds/actions/runs/39188328699. Original R437/full-T385 execution is https://github.com/Clear1nsight/integer-mult-bounds/actions/runs/39186343975. That workflow correctly failed its final exit-code gate on the T385 timeout. EXECUTION.json records exact sources and stage classifications. The later bounded replay is retained in the t385-bounded-replay artifact of the publication run. Both scientific template inputs are permanently retained in the package without foreign Git-history dependencies.

Remaining obligations and predecessor credit

A longer fresh T385 replay must complete discovery, both outer stages, and release. The all-size compiler, mutable-SOURCE, arbitrary capture-stream, whole-call hosting, precision/tape/recovery, analytic, and BIT interfaces remain conditional. This package does not re-prove PR3700's full word identity or admit either construction. Structured next steps are in TASK.json.

Credit ai-nanachi for PR3647's construction and actual templates; Sagar Thakrar for PR3700's old-head independent audit, OQ4 placement idea, and reused price engine; Ali Romman for PR1043's helper470 deletion; and the selected chain's credited authors in the retained notices. Credit dataisfire for PR3734, hexapode for PR3802, ChaloupeMusic for PR3815, and their native/owner/outer predecessors. New checker, controls, execution, and write-up by Chris R (@Clear1nsight), with substantial OpenAI assistance. No independent human or formal-kernel review is claimed. Apache-2.0 and upstream notices are retained.

Canonical main and both target PR heads were rechecked before publication and remained unchanged. Priority revision is 6. Live relevant PR searches and comments were read. Discussion coverage was unavailable, and an attempted coordination comment was denied with HTTP 403; no posted reservation is claimed.