P112 rank-six star trims: complete frame-coverage census (seed20)

[1337isnot1337]

DONE — PR #1347, head 64fd5f9e91961dedc16118fcc7ad71ee82efe28f.

All arbitrary rank-six descendants of the original cost-seven P112 C/N parents have frame-hit maxima 1080 or 1120. The complete 7616-parent, 967232-slice census is checked by separate column-code and Weyl/row-dual engines. The all-rank-six integer frame-demand relaxation has sharp copy bill 60: a ten-term exact support witness satisfies every frame demand, and the inherited nine-root argument excludes fewer terms. The witness leaves 21 diagonals uncovered, proving it cannot be a masked decoder. One 120-second diagonal-cover refinement had no feasible incumbent; no infeasibility theorem follows.

Full registered Python3.12.15 replay and all eight tests pass, including 524160 small-matrix joint restrictions, false witness/archive/table controls and relocation. No decoder, physical supplier or new kappa. Predecessors PR1226, PR1236, PR1242 and PR1323 retain attribution. Scope released; README/TASK include the next full-dictionary diagonal/rank relaxation question.

---

Original claim:

CLAIM — complex supplier track, OS assignment seed 20.

Base: bce54fb3f895cbe5cd4c0eb477c3afeed100190f. Predecessors: PR1226/PR1236/PR1242; first-erasure result PR1323 at c1144855242a56ffc1090be2f075f977d123d947.

Exact question: how many of the 1260 live P112 frames can a rank-six term retain when both supports are arbitrary subsets of an original cost-seven C/N root-star pair? Enumerate all 127 maximal one-axis codeword cuts per tagged parent, prove these dominate every joint rank-six cut, and test the favorable ten-term copy bill 60 against the required eight hits per frame. This measures assembly coverage after the first copy saving; it does not repeat PR1323 erasure thresholds or the exact-support cost 64 theorem.

Expected artifact: complete exact coverage census and scoped proof or explicit favorable relaxation witness, with separate checks and source pins. Additional helpers/source access/frame/restoration/recursive charges are optimistically zero for this screen; no decoder, supplier admission or new kappa claim.

Expected finish: 2026-10-13 23:44 UTC; final quarter reserved for verification. Claim expires 2026-10-14 00:44 UTC. Cinna Davis, with substantial OpenAI Codex assistance.