This tool assesses the accuracy of past predictions using metrics like the Brier and log scores, and it quantifies multi-source agreement to identify consensus probability and potential outlier sources.