We want to be able to re-run failed tests.
Our suggestion for a quick solution is to provide us an API to create "on the fly" new test pack with failed tests only and to re-run brand new test pack.