Fix broken or flaky tests

Give your AI agents the context to fix tests. Humans get the evidence.

Your engineers shouldn’t be babysitting CI