Binders were generated using MOSAIC, such that the target sequence is that provided by Adaptyv and the pipeline identical to that used in the Nipah Binder Competition. Binders were designed to be 80 amino acids long to minimize computational cost and reduce potential binding interfaces that could interfere with high-affinity binding. Generated binders were evaluated using the native MOSAIC loss function before being visually examined in Alphafold 3 and Boltz-2 predictions to confirm structural viability. Upon generating a favorable set of candidates, Scramble was used to optimize the binders using the known complex between RBX-1 and Glomullin. RBX-1 contacting residues on glomulin were trimmed before being inserted with candidate sequences into Scramble. For the purposes of this competition, the trunk module was not especially utilized (besides basic filtering) given the low number of variants produced. Regardless, a number of generated variants outperformed their parent binders in the MOSAIC ranking-loss assessment.