Kinases have proven valuable targets in successful cancer drug discovery projects, but not yet for malignant brain tumors where type-II inhibition of cyclin-dependent kinase 5 (CDK5) stabilizing the DFG-out inactive state has potential for design of selective and clinically efficient drug candidates. In the absence of crystallographic evidence for a CDK5 DFG-out inactive state protein−ligand complex, for the first time, a model was designed using metadynamics/molecular dynamics simulations. Glide docking of the ZINC15 biogenic database identified [pyrimidin-2-yl]amino-furo[3,2-b]-furyl-urea/amide hit chemical scaffolds. For four selected analogues (4, 27, 36, and 42), potent effects on glioblastoma cell viability in U87-MG, T98G, and U251-MG cell lines and patient-derived cultures were generally observed (IC 50 s ∼ 10−40 μM at 72 h). Selectivity profiling against 11 homologous kinases revealed multikinase inhibition (CDK2, CDK5, CDK9, and GSK-3α/β), most potent for GSK-3α in the nanomolar range (IC 50 s ∼ 0.23−0.98 μM). These compounds may therefore have diverse anticancer mechanisms of action and are of considerable interest for lead optimization.