It should be possible to fail a test based on the result of a JS execution