claude-council

Consult multiple AI coding agents (Gemini, OpenAI/Codex, Grok, Perplexity, plus the codex / gemini CLIs when installed) in parallel, then synthesize their answers side-by-side with consensus, divergence, and recommendations. Use when one model's bias could mislead a decision and cross-checking matters - architecture choices, debugging dead ends, security reviews, framework picks, tradeoff analyses. Supports specialized roles (security, performance, maintainability, devil, simplicity, scalability, dx, compliance), two-round debate mode, agent-enhanced deep analysis, quiet synthesis-only mode, auto-context file injection, response caching, and markdown export.