This enables a really cool workflow: 1. Someone reports a problem in Slack 2. Ship gathers context from the code and the running app 3. It then reproduces the failure and captures evidence 4. It investigates the likely cause of the problem. 5. It creates a ticket with all the information 6. An engineer or coding agent works on the fix 7. Ship creates a regression test to run it going forward