So that it would be possible to identify the test in the server log when the deployment fails.

Unlike in TCK there is no test archive builder/wrapper and so this will probably require quite a lot of manual work.