openai/gpt-chat-latest → A (9:1)

Side A removes a demonstrably dead zero-ratio early return from `GroupState::apply_vote`, relying on existing parser/browser validation and the existing `add_edge_weight` zero-weight check, then updates the test to verify the new intended behavior (items and pairs are still registered while zero-weight edges are skipped). Side B is an enormous seed commit mixing many unrelated additions, including files that accidentally contain shell transcript text (e.g. `forms.rs` and `ranking.rs` begin with terminal output), making it noisy and likely non-buildable rather than a clear, lasting project improvement.