The intent chapter's audit-loop table never iterates on an INCONCLUSIVE verdict and never rolls it up as UNACHIEVABLE; the build chapter's fix loop fails the round on INCONCLUSIVE exactly as on NOT_MET.

The intent chapter's audit-loop table never iterates on an INCONCLUSIVE verdict and never rolls it up as UNACHIEVABLE; the build chapter's fix loop fails the round on INCONCLUSIVE exactly as on NOT_MET.

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

• End A: .abcd/development/brief/04-surfaces/05-intent.md:592 — "Fail-closed: recorded as the shipped audit records it, no iterate, no summons, no replan; never flips to UNACHIEVABLE."
• End B: .abcd/development/brief/04-surfaces/34-build.md:712 — "a criterion it could not decide (INCONCLUSIVE) fails the round exactly as a NOT_MET one does"

05-intent's state-coverage table for the loop-to-acceptance audit loop says an INCONCLUSIVE rollup is fail-closed: no iterate, no replan, and it never flips to UNACHIEVABLE (the table is headed 'the loop trigger is NOT_MET only'). 34-build realises that same loop as the build lane's fix-round bound. There an INCONCLUSIVE criterion fails the round and goes to a fresh implementer, using up fix rounds. A lane out of rounds is handed back with the verdict unachievable, which lists the criteria the audit could not decide. One page says undecided criteria never drive re-work or an unachievable end; the other makes them do both. 05-intent says its per-intent shape predates the loop's re-scope into build, but its table still states the opposite policy.

Evidence: .abcd/work/reviews/2026-10-26-consistency/00-summary.md, finding 406.