bts-debatelisted

# Expert Debate Debate the given topic using expert personas. Read `.bts/config/settings.yaml` for `debate.expert_count` (default: 3) and `debate.rounds` (default: 3). ## Protocol You will play `debate.expert_count` expert roles sequentially. Each round, all experts speak. ### Setup Choose relevant expert perspectives for the topic. Example: - For "OAuth2 vs JWT": Security Expert, Performance Expert, Operations Expert - For "SQL vs NoSQL": Data Architect, Scale Engineer, Developer Experience Expert - For "Monolith vs Microservices": System Architect, DevOps Engineer, Team Lead ### Evidence standard (all rounds) Positions that rest on framework/platform behavior or on "X is the recommended approach" claims MUST carry a `Source:` line per `.claude/rules/bts-evidence-policy.md` (Context7 → official docs → site-filtered search; blogs and tutorials are not evidence). Look up disputed claims DURING the debate rather than arguing from memory — an uncited claim is an opinion, and /bts-adjudicate scores it as weak evidence. When the debate topic is an architecture choice, read the research doc's `## Official Guidance` section first; experts must engage with the vendor-recommended pattern, not ignore it. ### Round 1: Position Statement Each expert states their position with supporting evidence. ### Round 2: Rebuttal Each expert responds to the others' positions. Point out weaknesses, ask questions. ### Round 3: Synthesis Experts seek common ground. Propose a conclusion that a