Ingest market data, backtest a supplied set of candidate strategies with Qlib, and run the TradingAgents swarm debate to vet the backtested evidence as input to an allocation decision. Use when a researcher wants evidence-linked support for weighting multiple strategies; this workflow does not execute the allocation or place trades.