Multi-model deliberation as MCP tools, with an auto-triggering skill. Consult a panel of models (Claude, GPT, Gemini, Grok, Qwen, GLM, Kimi) on contested decisions where a single model might be confidently wrong, e.g. architecture choices, plan reviews, pricing tradeoffs. Each round returns per-model prose plus a cross-model claim map showing where the panel agrees and where it splits.