openai/gpt-chat-latest → B (20:1)

Side B fixes a substantive correctness bug in `compute_scores_from_edges` by switching the Rank Centrality normalization from summed edge weights to the canonical degree-based `d_max`, preventing oscillating bipartite Markov chains and producing correct rankings for star topologies. It also adds targeted Rust and end-to-end regression tests with ranking fixtures, whereas Side A is overwhelmingly rustfmt-driven reformatting plus minor tooling changes (pinning rustfmt/clippy, VS Code settings, and a small CSS cleanup) with little lasting behavioral impact.