[petalas]

Summary

Prove mixed-rank P112 shortening over every characteristic-zero field, then exclude the literal historical TWOSIDE (6,4) profile 5^5,6^5.

three endpoint ranks 5,5,6; one common support<=37
  code union<=81 forces an E6-normal A2 plane
  singleton count requires82 incidences against cap80

actual 5^5,6^5 caller, code dimension>=6
  delete four pieces; retain two code dimensions
  small5^3 support<=36; endpoints5,5,6
  contradiction84>80

Only research/p112-mixed-a2-caller/ is added. This is a sufficient threshold, with no sharpness claim. The proof avoids the invalid quotient premise 37+20<53 by classifying the full two-dimensional code union first.

Evidence

• Before: the source-pinned uniform endpoint theorems do not directly cover this caller's common support 36. The original PR3154 filter accepts the literal profile, with excess 35<36.
  After: a written mixed-rank theorem and explicit coefficient-deletion/restriction transfer exclude it. Both original and new filter implementations agree over all 7,546 cell multisets and reproduce 69 historical survivors. This is not a current unresolved-profile count.
• Registered CPython 3.12.15 replay and eight optimized behavioral groups pass. Replay checks all 1,120 A2 unions, three 2,667-space rank-five fiber censuses, three 127-hyperplane rank-six fiber censuses and all 112 original root/label pairs.
• A designated worker's read-only peer review found no blocking issue in the actual classification, count or caller transfer, matched all 16 reviewed files and all 6 retained Git sources, and passed optimized primary replay. Its unrun obligations are preserved in REVIEW.json. All reviewed bytes stayed unchanged; only that review receipt was added.
• Ruff E/F/B/I/UP/C4/SIM and formatting pass. All 19 committed package/input byte comparisons and archive integrity pass. The exact committed read-only Git-free export passes primary and all eight controls on optimized CPython 3.15.0.

python3.12 -W error -B scripts/research_package.py \
  --package research/p112-mixed-a2-caller

Commit cdeee42ba34d4da3db8c2b1fc9cbc134c7172fed; scientific and submission main 463113d675af7410ac91d53a4fed51ec6494eb9a, priority revision 6. Canonical investigation. Submission overlap covers 302 open PRs, 2630 closed/merged PRs and 2088 Discussion records with complete pagination and no warnings. The exact scope matched only this contribution's claim/package; this is not an exhaustive implication or novelty audit.

Credit DreamingOfClouds/PR3154 at 708f0ad8928c383de8294aea3c126e9461d2066d, Nick Petalas/PR3249 at 22c7bf316d55e59f381cc9ba833afccc7702fd58 and PR3255 at 2977a895c35fa019f53f181edee45177c77ec436, plus the retained root-simplex/E8/P112 lineage including joaqcoo. Global flat capacities 12/20, uniform reflection transport and the Weyl-chamber theorem remain inherited dependencies; the global flat censuses were not rerun. Open next questions identify the support38 classification gap and the need for actual caller reconciliation.

Merge Danger

Door: two-way

Blast Radius: isolated

One new research package. No frozen source, shared verifier or selected result changes. This establishes no full C5, complete decoder floor, physical supplier, native restoration, all-size multiplication theorem or new kappa. No independent human review, proof-kernel verification, Lean, Docker or selected cold replay is claimed.