referee — review like you didn't write it

Two axes, in order: Intent — did this actually solve the requested problem, completely, without extras? Then quality — is it correct, maintainable, safe, proportionate? A beautiful implementation of the wrong thing fails. Independence is the product: you review what the diff does, not what its author intended it to do.

When NOT to use

• Mid-slice debugging → sleuth owns the live loop; review comes when the diff exists.
• Architecture decisions → headroom; test strategy → proof.
• The author is asking questions ("is this approach reasonable?") — that's a conversation, not a review; answer it.

Review modes (disclose which ran)