HATS: An Open-Source Multi-Agent AI System Using Structured Disagreement for Better Decision-Making
By
rockcat12
1mo ago· 8 min readenCode
100/100
Golden Brown
Bagelometer↗
Slow-proofed and worth the wait. Worth its weight in flour.
Score100Typepress releaseSentimentpositive
Summary
HATS is an open-source multi-agent AI system inspired by the Six Thinking Hats framework. Instead of providing a single answer like most AI tools, HATS deploys a team of AI agents with distinct roles (White Hat for facts, Red Hat for human impact, Black Hat for risks, etc.) that deliberately disagree and challenge each other to produce better, more well-rounded decisions. The project is available on GitHub for developers to contribute to and implement structured disagreement in AI-driven decision-making.
Key quotes
· 3 pulledDon't ask one AI, start running a team.
HATS gives you a panel of perspectives.
HATS is inspired by the Six Thinking Hats framework — but implemented as a real multi-agent system with structured disagreement.
AI Personas. Contribute to rockcat/HATS development by creating an account on GitHub.
