When tests fail, and the user wants to configure them as known failing, it is a bit laborious. The user must copy all of the failing test case names (which are strewn across many lines, so not as simple as copying a region of output) and paste it into a file. If they want to consolidate them using wildcards (which is helpful so the file is smaller and more readable), it requires a little bit of mental math and analysis to determine the patterns.
Ideally, the tool itself would recommend the configuration to use and even compute the minimal representation using wildcards for you and include it in the output when there are failing test cases.