They mentioned that their automated monitoring did not pick up the issue and that they are working on improving the tooling around this.