Show GN: 31개의 LLM을 교차검증 체계로 엮은 100% 자율 주식 매매 시스템

This project demonstrates how multi-agent LLM systems can be applied to real-world financial trading with practical solutions to overcome AI limitations like hallucinations and cognitive biases. The architecture showcases sophisticated prompt engineering techniques (Bounded Autonomy, Red-Teaming) to mitigate the disposition effect and other systematic errors. By validating results across 31 specialized agents before executing real trades, the system provides a compelling case study for building trustworthy AI agents in high-stakes domains.