Every candidate gets filtered Every paper trade gets tracked Execution costs are measured on-chain Rugs are monitored The score is tested against actual outcomes

Current paper results: 53% win rate · +2.3% avg PnL · 0 confirmed rugs And here's the part we're most interested in: The data says 80% of our current score weighting isn't proving alpha. So we're not pretending it works. We're testing what does.