Claude Code Council + LLM Council 🤖

This repository contains two powerful AI systems:

1. Claude Code Council - Multi-agent code review system
2. LLM Council - Multi-model AI consensus system ✨

---

🆕 LLM Council - Multi-Model Consensus System

Get validated answers from Claude, GPT-4, Gemini, and Grok simultaneously!

Instead of asking each model separately, ask them all at once and get:

• ✅ Consensus analysis
• ⚖️ Weighted voting
• 🔍 Meta-analysis of disagreements
• 🎯 Synthesized final answer

⚡ Quick Install

# One command installation
bash install.sh

🚀 Usage

# Ask a question
llm-council ask "What is the best way to learn Python?"

# Or short syntax
council "Explain async/await"

# List available models
llm-council models

# Setup API keys
llm-council setup

🔑 Authentication - NEW: CLI Session Support!

Use your existing CLI subscriptions instead of API keys! Save costs by leveraging Claude Code CLI, ChatGPT CLI, and other tools you already have installed.

Automatic CLI Authentication:

# No API keys needed if you have CLI tools installed and authenticated!
# LLM Council will automatically use:
# - Claude Code CLI session
# - ChatGPT CLI (shell_gpt) session
# - Gemini via gcloud CLI
# - Falls back to API keys if CLI auth unavailable

Supported CLI Tools:

• Claude Code CLI (claude)
• Anthropic CLI (anthropic)
• ChatGPT CLI (chatgpt, sgpt)
• OpenAI CLI (openai)
• Gemini CLI (gemini)
• Google Cloud CLI (gcloud)

Priority: CLI Session → API Keys (automatic fallback)

See CLI_AUTH.md for detailed setup and usage.

📚 Documentation

• Main README - Full documentation
• Quick Start - Fast setup guide
• Update Guide - Development workflow
• Install Guide - Pipx installation details
• Examples - 18 usage examples
• Changelog - Version history

🔄 After Each Code Change

pipx reinstall llm_council && llm-council version

That's it! Changes applied.

---

📦 Claude Code Council (Original)

AI-powered multi-agent code review system for Claude Code.

See original README below for details.

---

Claude Code Council (Original Project)

AI-powered multi-agent code review system for Claude Code

Get intelligent, evidence-based code reviews from 5 specialized AI agents working in parallel. Catch security issues, enforce DRY principles, protect your APIs, and maintain architectural consistency—all before you commit.

[Continue reading original README...]