Research: seed 5 exact joint-boundary census for PR589 r659 (no kappa)

[1337isnot1337]

Result

Adds a scoped negative-result audit for seed 5: the exact jointly boundary-adaptive frame census for PR589 A[2071] and B[820] at shared register 659. Out of 20 dynamically generated frame assignments, 12 pass both reciprocal boundary checks and complete touched-register chains; the remaining 8 fail the restored A upper boundary. The legal assignments yield 8 paid-profile ties and 4 strict losses, with no improvement or unresolved comparison. The fixed-original-boundary control also yields the same 12 legal pairs, so this cut exposes no joint-only choice.

All 12 legal assignments replay through wlib.rebuild_C and wlib.paid. The baseline profile, relaxed-boundary rejection, exact rational loss witnesses, frame domains and source pins are recorded in the package. This does not claim a stronger bound or a new kappa.

Reproduce

Base commit: bce54fb3f895cbe5cd4c0eb477c3afeed100190f.

python3 research/bit589-boundary-adaptive-r659-seed05/verify.py

Expected: PASS_BIT589_BOUNDARY_ADAPTIVE_R659 seed=5 candidates=20 valid=12 joint_only=0.

The package metadata/source pin check passes. Local direct verification ran on CPython 3.14.7; the registered replay requires Python 3.12, which is unavailable locally, so its hosted check is the remaining interpreter-specific verification.

Scope and credit

This tests one seed-selected operation pair on the pinned PR589 word. It does not cover other registers, larger move sets, other words, or recursive constructions. The accepted #1094 selected result and its assumptions are unchanged; the assembly is still complex-limited.

The task selection records three candidates and the OS-random seed. This work extends the fixed-option census/open question in #1382 using the exact transport and pricing methods of #1345. Prepared by Cinna Davis with substantial OpenAI Codex assistance. New package material is Apache-2.0.

[Mister-Meeseeks]

100-PR batch review at 3854fbf15402cccd6758f839c93e824df3041ba9.

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

Reviewed scope: Jointly boundary-adaptive A/B transport census for one specified PR589 register659 pair: all 20 candidates, 12 legal and all literal profiles replayed. No improvement in that pair; no general fixed-option completeness claim.

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.