councillisted

5-model strategic advisory. Ask the same question to five different LLM seats in parallel, save each response to a persistent file, then synthesize a recommendation. Each seat is locked to a role (Theorist, Validator, Provocateur, Diagnostician, Field Engineer) so the responses disagree by design instead of converging by accident. Use when the user says "ask the council", "get opinions", "council on", "what do they all think", "multi-model opinion", or wants strategic input from multiple LLMs before making a decision.
eprouveze/claude-skills · ★ 0 · AI & Automation · score 78

Install: claude install-skill eprouveze/claude-skills

# Council — 5-Model Strategic Advisory Ask the same question to five LLM seats, each locked to a thinking role, then have the orchestrator (Claude) synthesize. Quality matters more than latency or cost (~$0.50–$1.00 per call, 60–180s wall time). ## The five seats | Seat | Default model | Role | | --------------- | --------------------------- | -------------------------------------------------------------------- | | Theorist | Gemini 3 Pro (or 2.5 Pro) | Research-grounded, framework-driven, lateral alternatives | | Validator | gpt-5.5 + xhigh reasoning | Eval methodology, defensible cases, instruction-following | | Provocateur | Moonshot Kimi K2.6 | Challenges premises, surfaces uncomfortable truths | | Diagnostician | gpt-5.5 + xhigh reasoning | Names the mechanism, cites stratified evidence, commits to action | | Field Engineer | DeepSeek-v4 (R1 reasoning) | Ready-to-paste artifacts, tabular root-cause analysis, iterative | The orchestrator (whichever model is running the skill) does the Blue Hat synthesis pass. Including the orchestrator as a seat would be self-querying. Each seat is overridable via environment variables — see "Configuration" below. ## Preparation protocol The brief IS the work. Budget 15–30 minutes assembling it before launching any seats. 1. Research the topic. If your stack has