What would make a non-blocking shadow pilot worth your time?

[daichunghy]

I do not want a PatchGate pilot to start with "install this and tell me what you think." A useful shadow run should answer a smaller question: did the result help a maintainer decide what to look at, or did it add another noisy check?\n\nThe current pilot idea is deliberately non-blocking. It should not change merge eligibility, request repository secrets, or execute pull-request code in the privileged metadata lane. The feedback I need is concrete: false blocks, unknown causes, unclear remediation, or a result that arrived too late to help.\n\nIf you have tried a similar shadow workflow, what did you measure before deciding whether to keep it?