Llm Council

This is Karpathy's LLM Council pattern made provider-agnostic. Three phases: models answer independently, rank each other's anonymized responses, then a chairman synthesizes everything into a final answer. Works with OpenAI, Anthropic, Fireworks, OpenRouter, or any OpenAI-compatible endpoint you configure via env vars. The --wiki flag persists the full transcript as a searchable page, which is genuinely useful for architecture decisions you'll want to reference later. Cost scales quadratically with council size since every model ranks every other response, so stick to 3-5 models unless you have budget to burn. Best for high-stakes plan reviews or when you legitimately want multiple AI perspectives on record, not just a single model's output.

install copies · 90 days. Times the install command was copied on this site in the last 90 days, bots filtered.: 13
skills.sh installs: 175
GitHub stars: 2.9K
last updated: Jun 8, 2026

View on GitHub

Install

$ npx -y skills add rohitg00/pro-workflow --skill llm-council --agent claude-code

Installs into .claude/skills of the current project.

Files

LLM Council

Karpathy's LLM Council pattern with Anthropic and OpenAI-compatible providers.

When to use

High-stakes plan review (/plan crosses N-file threshold)
Conflicting learning-rules → re-resolve via vote
User invokes /council "<query>" or /wiki council
Architecture decisions where you want multiple viewpoints captured
Persisting deliberation as a wiki page for future reference

Three phases

Independent: each model answers in parallel
Ranking: each model ranks anonymized peer responses
Synthesis: chairman model reads all responses + rankings → final answer

Provider config

Configure optional keys in the plugin configuration dialog. Use the providers MCP server's run_provider_task tool with task: "council" and args: ["run", "<query>", "--provider", "openai"]. For provider status, pass args: ["providers"]. Never request keys in chat or read them from existing credentials or shell configuration.

SKILL.md on GitHub