Where our code gates actually spent their time
A per-edit hook went from 210ms to 15ms and a lint pass from 21s to 0.7s. The hooks spent their time starting processes, not checking code, and CI was bounded by its runners, not its tests.
A per-edit hook went from 210ms to 15ms and a lint pass from 21s to 0.7s. The hooks spent their time starting processes, not checking code, and CI was bounded by its runners, not its tests.