Some quantum measurements are complementary in a precise way. If a system has a definite outcome in one measurement basis, every outcome in another is equally likely. Mathematicians call these mutually unbiased bases. In six dimensions, three such bases are known. Whether a fourth can exist remains an open problem. This research candidate closes one particular route: a continuous family of matrices introduced by Szöllősi. Its proposed exclusion covers the whole specified two-parameter family, including boundary cases where numerical methods become difficult. It does not settle the general six-dimensional question. The key idea is to separate discovery from checking. One program proposes small regions containing possible vectors. Another program uses interval arithmetic to check that no relevant vector has been missed. It then turns possible relationships between the regions into a graph. Two extra bases would require a particular arrangement of twelve vectors. Ruling out that arrangement rules out a quartet. The method does not need to isolate every individual solution, which helps near degenerate parameters. The package also revisits the already known Fourier-family exclusion. That is a check of an existing theorem, not a second new result. Review led to explicit repairs concerning shared grid boundaries and stale cached verification results. The revised workflow binds each receipt to its actual inputs and checking environment. The work remains an unrefereed candidate, not a proof-assistant formalisation or independently established human peer review. This is Evidence Press, Szöllősi-family MUB exclusion, dated the thirtieth of September twenty twenty-six. The paper and evidence package are linked on the release page. This synthetic AI voice explains the research; it is not additional scientific evidence.