The arguments for creating the completeness evaluator.
Optionalchoices?: ClassificationChoicesMapOptionalinputMapping?: ObjectMapping<RecordType>The mapping of the input to evaluate to the shape that the evaluator expects
Optionalname?: stringOptionaloptimizationDirection?: OptimizationDirectionOptionalpromptTemplate?: PromptTemplateOptionaltelemetry?: TelemetryConfigAn evaluator function that takes a CompletenessEvaluationRecord and returns a classification result
indicating whether the conversation is complete or incomplete.
const evaluator = createCompletenessEvaluator({ model: openai.decisionModel("gpt-6-luna") });
const result = await evaluator.evaluate({
conversation:
"User: Reset my password and update the billing address.\nAssistant: Your password has been reset.",
});
console.log(result.label); // "incomplete"
Creates a completeness evaluator.
This function returns an evaluator that checks whether every active user request in a conversation was actually completed, not merely acknowledged.