First, let's look at the individual test that you are running from the master test. The individual test defines its own behavior. When there is a failure, you configure that test itself about whether it should abort or not when there is a failure. If the test itself needs to clean up after itself, then you can also configure an event handler to call a clean up procedure if you want.
Now let's look at the master test case. The EXEC.run step that runs the individual test produces a response that provides information on how that test ran. You can add analysis to that step if you want to take certain actions depending on the outcome of that test. For example, on a failure, you can call a cleanup method before moving on to the next test. If you want that same rule to apply to all of your EXEC.run steps, you can add a global rule to do the same thing.
AbortExecution and AbortTest actions are designed for controlling the currently running test. You may want these in the individual test, but it sounds like you do not want these in the master test. In the master test, it sounds like you want to use analysis on the EXEC.run steps to take special actions depending on the outcome.
