Being able to mark individual tests as flaky is really nice as we don't have to repeat the whole build job any more now! It also makes it easier to spot the flaky tests so we can inspect them to see whether we can make them "less flaky".

The only downside is of course that we'll now get some failed builds where we didn't mark tests as flaky yet. I just restarted some jobs that failed due to a single test and then created issues for these flaky tests (as all jobs were successful with the second run).

We should probably add the flaky marking to these tests shortly so they don't let our build fail for long.