Research audit: resolve R/R0 revert expectations with exact PR626 accounting

[thakrarsagar]

The selected PR626 proof note says its R0 revert controls should pass at R=492, while its own table and evidence correctly report failures when provider R replaces parent R0. This research audit resolves that contradiction with reproducible finite evidence. The selected implementation distinguishes the counts correctly; no new kappa or construction defect is claimed.

The package independently recounts the selected invoice from pinned native dimensions, dispatch, and provider-local receipts. It reproduces the complete parent-envelope price 1833791190783112, the rebase correction -300156, and the retained allowance total 4711738180507680. It also replays the original accounting slice and its nine finite guards, rejects seven one-site source changes and four altered inputs, and accepts the equal-literal-504 control identically.

Changing route R to 504 overcharges by 358516260864. Its rejection means disagreement with this exact receipt, not an unsafe conservative bound. Replacing parent-bank R0 by 492 undercharges the fixed envelope by 22407266304.

Reproduce with the Python standard library:

python3 -I -B research/r492-parent-accounting-audit/verify.py

Expected: PASS_R492_PARENT_ACCOUNTING_AUDIT: 9 finite guards, 7 source rejections, 1 identical control, 4 input rejections.

Validation passed under normal and optimized Python, with byte-identical evidence, and from a committed-file export in an unrelated working directory. A changed pinned source is rejected before replay. make selected-record-check passes. The full suite and cold supplier reconstruction were not rerun; all five changed files belong to this new audit package.

Base: 66849e1475c930fd7732b1a0067ffee7e1e95f60. Selected PR626 source: 660de893d27bb7abd1d5b01b1640207528a5c851. SOURCES.json pins all nine consumed inputs; README.md records the dependency heads and inherited assumptions. The original 32-site scratch suite and its R=504 fixture remain unreproduced. Full local algebra/admission, moment and outer constraints, and all-size/physical/analytic interfaces remain inherited. The selected frontier and upstream/ are unchanged.

Seed 70, adversarial mathematical audit. No pinned sprint coordination issue exists, so the claim is local. The refreshed queue through #722 contains no duplicate: #675 concerns owner timing and bridge omission; #717 concerns moment sensitivity prices.

Credit to Chafik Boukhalfa, Jeremy Kothe, GlobalR/AtomReviewer, icekylinx, DreamingOfClouds, Rohan Arun, eumemic and the retained lineage. Prepared for thakrarsagar with substantial OpenAI Codex assistance, including a separate agent review. Apache-2.0; this is not independent human peer review.

[Mister-Meeseeks]

100-PR batch review at 624ccfe44b1c5816f0279481d91cca6a5a2290ab.

Merged to main, preserving the original contributor commits and package attribution.

Reviewed scope: Historical R492 versus R0 accounting audit: source equations and invoices distinguished; no new frontier.

The pinned scoped command passed; the batch ledger contains its command, input hashes, environment and log. This incorporates the research/tooling contribution and does not change the selected κ or discharge the inherited multiplication-theorem assumptions.

Batch review, attribution and evidence. Thank you for the contribution.