Ingest market data, backtest a supplied strategy to surface its maximum-drawdown episodes, and run the TradingAgents swarm debate to vet a candidate recovery hypothesis for the observed drawdown. Use when a researcher wants evidence-linked drawdown analysis and a vetted recovery hypothesis; this workflow does not place trades or guarantee recovery.