Retry a job based on a particular output pattern
Sometimes we can identify a particular error based on the message that retrying can solve the problem, but we might not be able to detect and act upon this error in an easy way.
For example, this error: #58882 (closed)
We can rescue this and retry it in the test level, but doing this based on the error output can be more flexible because not all errors can be handled in this way.
retry: content: /if-we-match/
This is similar to #25738
Idea from @ayufan