3omarkeshk3

**One prompt, many frontier models, one synthesized answer.** Council AI fans your prompt across 25+ models from 9 labs (Claude, GPT, Gemini, Grok, DeepSeek, Qwen, Mistral, Kimi, GLM) in parallel, then a moderator model synthesizes a consensus answer with a numeric agreement score. **Tools:** `council_query` (parallel fan-out + consensus), `council_review` (verdict-first multi-model code review — SHIP/NO-SHIP, findings confirmed by 2+ models, attributed dissents — built for coding agents), `cou