What problem does it solve? Single-model planning can miss alternative approaches and blind spots. This Skill consults external AI models (ChatGPT and Gemini) on a question, then synthesizes their perspectives with Claude's own analysis into one implementation plan. ## Core Features & Use Cases - Multi-LLM Querying: Sends a prompt to both OpenAI and Gemini via CLI tools (codex, gemini-cli) with automatic API fallback. - Perspective Synthesis: Analyzes each model's suggestions, identifies valuable insights and concerns, and merges them into a final plan with per-model attribution. - Configurable Models: Supports model selection via environment variables, from cost-efficient defaults (gpt-5-nano, gemini-3-flash-preview) to premium reasoning models. - Use Case: Before architecting a real-time IoT data pipeline, ask the council for perspectives — ChatGPT might suggest Kafka while Gemini recommends edge computing, and the final plan incorporates both. ## Quick Start Ask Claude to consult the LLM council about how to architect your system, after adding your OPENAI_API_KEY and GEMINI_API_KEY to a .env file.
