BigBrain

概要

BigBrain functions as an MCP server that empowers Claude Code to interact with OpenAI's Codex and Google's Gemini CLIs. Designed to keep Claude in charge, it enables the primary AI to solicit 'second opinions' or parallel processing from other models. It achieves this by spawning the CLIs as asynchronous subprocesses, providing structured responses, and prepending project context like `CLAUDE.md` and `MEMORY.md` to every prompt sent to Codex or Gemini. This integration enhances Claude's problem-solving by allowing it to use various tools like asking specific models, querying both in parallel, requesting consensus, or initiating multi-round debates.

主な機能

• Integrates Claude Code with Codex and Gemini CLIs
• Spawns CLI interactions as async subprocesses for efficient multi-model querying
• Provides current project context (`CLAUDE.md`, `MEMORY.md`) to external models
• Offers tools for parallel querying, consensus requests, and multi-round debates
• Authenticates via cached OAuth for both Codex and Gemini CLIs
• 0 GitHub stars

ユースケース

• Obtaining second opinions on code or logic from multiple AI models
• Synthesizing agreements and differences in responses from various language models
• Facilitating multi-round AI debates to explore complex problems or solutions
• Leveraging different AI model strengths within a single development workflow
• Enhancing Claude Code's capabilities with external model insights for coding tasks