except that for the test usecase, we need metadata...
# development
w
except that for the test usecase, we need metadata back around how many attempts we made in order to determine that something was flaky
thoughts on this? my feeling is that the test goal actually needs high level retry to do the right thing with flaky tests.