Follow exporter recording options for classifier data

Author: andreiborza

packages/server-utils/src/ai/mastra/classifier-evaluation.ts

@@ -8,6 +8,9 @@ import type { Span, SpanTimeInput } from '@sentry/core';
  */
 export interface ClassifierEvaluationCall {
   span?: Span;
+  /** The recording options of the exporter that opened the span, so the call's data follows them. */
+  recordInputs?: boolean;
+  recordOutputs?: boolean;
   /** Mastra's end time, set when Mastra ended its span before the call settled. */
   endTime?: SpanTimeInput;
   settled?: boolean;

packages/server-utils/src/ai/mastra/index.ts

@@ -153,7 +153,8 @@ export class SentryMastraExporter implements MastraObservabilityExporter {
 
     const evaluation = span.type === 'classifier_evaluation' ? takeStartingClassifierEvaluation() : undefined;
     if (evaluation) {
-      evaluation.span = sentrySpan;
+      const { recordInputs, recordOutputs } = resolveAIRecordingOptions(this._options);
+      Object.assign(evaluation, { span: sentrySpan, recordInputs, recordOutputs });
     }
 
     this._trackSpan(span.id, { span: sentrySpan, spanType: span.type, usage: {}, evaluation });

packages/server-utils/src/integrations/mastra-classifier.ts

@@ -1,7 +1,5 @@
 import { GEN_AI_INPUT_MESSAGES, GEN_AI_OUTPUT_MESSAGES } from '@sentry/conventions/attributes';
 import { isObjectLike } from '@sentry/core';
-import { resolveAIRecordingOptions } from '../ai/core/utils';
-import type { GenAiOptions } from '../ai/core/utils';
 import type { ClassifierEvaluationCall } from '../ai/mastra/classifier-evaluation';
 import { setStartingClassifierEvaluation } from '../ai/mastra/classifier-evaluation';
 import { getEvaluationInputMessages, getEvaluationOutputMessages } from '../ai/typesafe';
@@ -20,7 +18,7 @@ interface ClassifierEvaluateChannelContext {
  * Add the evaluated state, questions and answers of a `Classifier.evaluate()` call to the exporter's
  * `classifier_evaluation` span, which Mastra leaves without input or output.
  */
-export function recordClassifierEvaluations(options: GenAiOptions): void {
+export function recordClassifierEvaluations(): void {
   const channel = diagnosticsChannel.tracingChannel<ClassifierEvaluateChannelContext>(
     CHANNELS.MASTRA_CLASSIFIER_EVALUATE,
   );
@@ -41,24 +39,19 @@ export function recordClassifierEvaluations(options: GenAiOptions): void {
     safeChannelCallback(() => {
       const call = calls.get(message as object);
       if (call) {
-        finishClassifierEvaluation(call, message as ClassifierEvaluateChannelContext, options);
+        finishClassifierEvaluation(call, message as ClassifierEvaluateChannelContext);
       }
     });
   });
 }
 
-function finishClassifierEvaluation(
-  call: ClassifierEvaluationCall,
-  message: ClassifierEvaluateChannelContext,
-  options: GenAiOptions,
-): void {
+function finishClassifierEvaluation(call: ClassifierEvaluationCall, message: ClassifierEvaluateChannelContext): void {
   call.settled = true;
-  const { span } = call;
+  const { span, recordInputs, recordOutputs } = call;
   if (!span) {
     return;
   }
 
-  const { recordInputs, recordOutputs } = resolveAIRecordingOptions(options);
   if (recordInputs) {
     const params = isObjectLike(message.arguments[0]) ? message.arguments[0] : {};
     // Questions given to the constructor take precedence over the ones passed to `evaluate()`, as in Mastra.

packages/server-utils/src/integrations/mastra.ts

@@ -97,7 +97,7 @@ function instrumentExporter(options: MastraOptions): void {
   });
 
   captureExecuteWithContextErrors();
-  recordClassifierEvaluations(options);
+  recordClassifierEvaluations();
 }
 
 /**

packages/server-utils/test/ai/lib/tracing/mastra-exporter.test.ts

@@ -190,6 +190,16 @@ describe('SentryMastraExporter', () => {
     expect(spanToStaticSpanJSON(call.span!).description).toBe('evaluate jev-1');
   });
 
+  it('passes its own recording options to the classifier evaluation call', async () => {
+    exporter = new SentryMastraExporter({ recordInputs: false, recordOutputs: false });
+    const call: ClassifierEvaluationCall = {};
+
+    setStartingClassifierEvaluation(call);
+    await run(started(makeSpan({ id: 'eval-1', type: 'classifier_evaluation' })));
+
+    expect(call).toMatchObject({ recordInputs: false, recordOutputs: false });
+  });
+
   it('records the agent-level prompt and response as gen_ai messages', async () => {
     const span = makeSpan({
       entityName: 'agent',