Gentrace Integration for Evaluation

Integrate Gentrace into your agent workflow. After an argument is generated, log the input prompt, the generated argument, and any intermediate steps from the Claude Agents SDK to Gentrace. Configure custom metrics in Gentrace (e.g., 'LogicalCoherence_Score', 'Persuasiveness_Rating') and implement a simple function to calculate these for initial evaluation. The goal is to track and improve the argument quality over iterations.