A reproduced numerical lower bound of 119 for the flexible job-shop benchmark behnke7

This research note reports a reproduced numerical lower bound of 119 for the flexible job-shop benchmark behnke7 (sm02_2; Sm02.2). The checked ScheduleOpt snapshot lists a lower bound of 117 and an upper bound of 124. A makespan-at-most-118 model returned INFEASIBLE with OR-Tools CP-SAT 9.15.6755. In a separate execution environment, an exact 180-second replay returned UNKNOWN; an otherwise unchanged replay with a 600-second limit returned INFEASIBLE after 438.10 seconds. The supplement includes pinned inputs, an exported model, code, actual execution logs, and a separate static checker. This is numerical evidence from the same solver implementation, without an externally checked formal proof certificate. It does not establish an optimal makespan of 119, a new algorithm, global priority, or independent validation of the archive's upper bound. The note has not been peer reviewed. Byungwoong Yoo directed the research and verification workflow; AI tools assisted implementation, execution, source checking and preparation. The note is licensed CC BY 4.0; identified Python code is MIT-licensed. Included benchmark data retain their original CC BY-SA 4.0 or MIT terms, as documented in the supplement.