The mental model describes the capture-kind classifier as operating; the naming register marks it staged with no classifier sub-verb registered

The mental model describes the capture-kind classifier as operating; the naming register marks it staged with no classifier sub-verb registered

A premise contradiction (severity minor) found by the cross-document consistency pass.

• End A: .abcd/development/brief/01-product/03-mental-model.md:74 — "The capture-time classifier (intent classify-capture-kind, itd-44) recognises this signature and emits a fourth verdict, decision."
• End B: .abcd/development/brief/02-constraints/04-naming.md:184 — "Capture-TIME classifier verdict, produced by the capture-kind classifier the intent surface is designed to gain; no classifier sub-verb is registered today."

The mental model says, without a staged marker, that the capture-time classifier intent classify-capture-kind recognises the signature and emits the decision verdict. The naming register marks the capture verdict staged and says no classifier sub-verb is registered today. The truth rule needs a staged claim to say so, and here the two pages give the classifier different statuses.

Evidence: .abcd/work/reviews/2026-10-26-consistency/00-summary.md, finding 13.