Package dev.braintrust.eval

Interface TracedClassifier<INPUT,OUTPUT>

Type Parameters:

INPUT - type of the input data

OUTPUT - type of the output data

All Superinterfaces:

Classifier<INPUT,OUTPUT>

public interface TracedClassifier<INPUT,OUTPUT>
extends Classifier<INPUT,OUTPUT>

A classifier that receives access to the full distributed trace of the task that was evaluated.

Implement this interface when your classifier needs to examine intermediate LLM calls, tool invocations, or other spans produced during task execution — not just the final TaskResult.

Field Summary

Fields inherited from interface dev.braintrust.eval.Classifier

INVALID_CLASSIFICATION_MESSAGE

Method Summary

Modifier and Type	Method	Description
default List<Classification>	classify(TaskResult<INPUT,OUTPUT> taskResult)	Classifies the result of a successful task execution.
List<Classification>	classify(TaskResult<INPUT,OUTPUT> taskResult, BrainstoreTrace trace)	Classifies the task result using the distributed trace for additional context.

Methods inherited from interface dev.braintrust.eval.Classifier

getName

Method Details

classify

List<Classification> classify(TaskResult<INPUT,OUTPUT> taskResult,
 BrainstoreTrace trace)

Classifies the task result using the distributed trace for additional context. Called instead of Classifier.classify(TaskResult) when a BrainstoreTrace is available.

Parameters:

taskResult - the task output and originating dataset case

trace - lazy access to the distributed trace spans for this eval case

Returns:

zero or more classifications

classify

default List<Classification> classify(TaskResult<INPUT,OUTPUT> taskResult)

Classifies the result of a successful task execution.

When used inside an Eval, this overload is never called — classify(TaskResult, BrainstoreTrace) is dispatched instead. This default implementation throws UnsupportedOperationException to surface any accidental direct calls.

Specified by:

classify in interface Classifier<INPUT,OUTPUT>

Parameters:

taskResult - the task output and originating dataset case

Returns:

zero or more classifications. An empty list means "no classifications for this case".