openai/gpt-chat-latest → A (3:1)

Side A introduces substantive security and correctness improvements: voting now fails closed instead of silently falling back to an anonymous actor, OAuth mocking is gated behind an explicit environment flag, Secure cookie handling is centralized, open-redirect sanitization is strengthened, and these behaviors are covered with new tests. Side B adds useful functionality around a pairwise voting UI, pair selection logic, and ID normalization, but much of the patch is feature work and refactoring rather than the broad, lasting hardening of core authentication and session handling delivered by Side A.