The pattern in legal AI incidents has been institutional contrition followed by vendor reassurance — neither of which produces anything a practitioner can use. cite.review breaks that pattern by converting Sullivan & Cromwell's public failure directly into open-source tooling [1]. An apology to a federal judge does not prevent the next filing; a free, auditable verification layer does. The practitioners already inside the 30-day windows that imposes on court AI policy are not waiting for the next firm to apologize — they are already…