与其赌单一模型的一次判断，不如让多个智能体各自分析、再投票形成共识。Rather than betting on one model’s single call, let several agents analyze independently and vote into a consensus.
这是一个用多智能体共识做决策的加密量化回测框架，强调透明规则与纪律——这里只展示方法论，不涉及任何具体策略或参数。它是我"技术复杂度"的一个证明：能把一件普通人以为很难的事做出来。This is a crypto quant backtesting framework built on multi-agent consensus, with transparent rules and discipline — shown here as method only, no strategy or parameters. It is a proof of technical depth: doing something most people assume is hard.