Automated notification when CI tests fail repeatedly in rolling builds

It was suggested that we could use RunFo to determine when a particular test had enough failures that it should be reported.

This could be a pipeline run on a schedule. Need to determine how many days of rolling builds to take into account and what the threshold is for reporting.

Start by investigating how RunFo exposes repeated test failures across rolling builds and how a scheduled pipeline could consume that data. The open decisions are the rolling-build time window and reporting threshold; done means an agreed, scheduled notification reports tests that exceed the threshold.

Written by the indexing model from the issue text.