The exact fourth-vector feasibility wall of the canonical MUB triple in dimension six

This record contains a paper and runnable verification bundle proving an exact fourth-vector feasibility wall for the canonical product MUB triple in dimension six.

***The result does not solve MUB(6)****

It quantifies the pinned canonical triple exactly:

The exact value is max_v f(v) = (88 + 3*sqrt(6)) / 100, approximately 0.953484692283495.

where $f(v)=1$ exactly characterizes a vector unbiased to all three bases of the triple. The matching global upper bound is certified by an exact PSD/Veronese certificate and deterministic symmetry closure over the triple’s 432-element symmetry group, with no randomized or floating-point step in the proof path.

Start with 00_README.pdf, which is set as the preview/front door. It states the scope, theorem, non-claims, quick verification commands, and useful community checks. The deposit also includes the full manuscript, the complete verification bundle, README.md, VERIFYING.md, SHA256SUMS, and citation/license metadata.

Core verification path:

sha256sum -c SHA256SUMS
make verify-global
python3 python/mub6_d6_wall_attainment_exact.py
python3 python/mub6_d6_wall_local_certificate.py
cd lean && lake build Mub6Lemmas

The invitation is not “believe this.” The invitation is: run it, break it, route it, and help map the rest of the MUB6 wall.

Companion community project: 

MUB6 Wall Atlas,  a public, nonfinancial atlas/game layer for organizing independent reruns, exact bricks, Qiskit teaching puzzles, Lean/formalization tasks, and next-wall candidates. The atlas is not part of the theorem proof path; it is a routing and collaboration layer for mapping the rest of the wall.

https://github.com/invariant-systems-ai/mub6-wall-atlas