HATS is a multi-agent AI platform that utilizes the Six Thinking Hats framework to facilitate structured debates, project management, and collaborative decision-making through animated 3D avatars.

Key Points

• Employs six distinct AI personas—White, Red, Black, Yellow, Green, and Blue—to simulate diverse perspectives and challenge user assumptions.
• Integrates a Kanban board for task management, allowing agents to automatically update, move, and resolve tickets via the Model Context Protocol.
• Supports multiple LLM providers including OpenAI, Claude, Gemini, and local models via Ollama or LM Studio.
• Features real-time 3D avatar rendering using Three.js, with synchronized speech and lip-syncing powered by Piper TTS and Rhubarb.
• Offers five specific meeting types, including standups and sprint planning, with full transcript generation and project isolation.

Why it Matters

By replacing standard single-response AI interactions with a debate-oriented multi-agent system, HATS helps teams identify blind spots and mitigate confirmation bias. This structured approach to AI collaboration provides a more rigorous framework for complex decision-making and project execution in professional environments.