A red pipeline is not useful feedback on its own. The developer still has to discover what failed, whether it is their change, and what to do next.
Optimize for the failure path
The happy path should be quiet. The failure path deserves the design work: specific job names, a short summary, and the most relevant diagnostic close to the top.
Make ownership visible
When an infrastructure issue blocks a build, the message should say who owns the system and where its current status lives. Good feedback turns interruption into direction.