Most coding agent benchmarks skip large-scale refactoring. Not this one.
The New Stack's newsletter signup page highlights that most coding agent benchmarks skip large-scale refactoring, introducing a new benchmark that addresses this gap. The page invites software engineering leaders to subs…