launchdarkly · kinyoklion · Aug 19, 2026
@@ -69,6 +69,24 @@ There are several attributes with special handling within a single or multi-cont
 - `anonymous` - Must be a boolean. Equivalent to `anonymous` in the SDK.
 - `name` - Must be a string. Equivalent to `name` in the SDK.
 
+### Flag Metadata
+
+Evaluation results include LaunchDarkly specific information in the OpenFeature flag metadata. A key is only present when it applies to the evaluation.
+
+| Key | Type | Description |
+|---|---|---|
+| `variationIndex` | number | The index of the served variation. Absent when the default value was returned. |
+| `inExperiment` | boolean | Present, and `true`, when the evaluation was part of an experiment. |
+| `ruleIndex` | number | The index of the matched rule, for a `RULE_MATCH` reason. |
+| `ruleId` | string | The identifier of the matched rule, for a `RULE_MATCH` reason. |
+| `prerequisiteKey` | string | The key of the failed prerequisite flag, for a `PREREQUISITE_FAILED` reason. |
+| `bigSegmentsStatus` | string | The status of the Big Segment query, when the evaluation required one. |
+
+```typescript
+const details = await client.getBooleanDetails('my-boolean-flag', false);
+const inExperiment = details.flagMetadata.inExperiment === true;
+```
+
 ### Examples
 
 #### A single user context

@@ -83,7 +83,12 @@ it('resolveBooleanEvaluation returns the translated result for boolean flags', a
     { kind: 'user', key: 'u' },
     false,
   );
-  expect(result).toEqual({ value: true, variant: '1', reason: 'OFF' });
+  expect(result).toEqual({
+    value: true,
+    variant: '1',
+    reason: 'OFF',
+    flagMetadata: { variationIndex: 1 },
+  });
 });
 
 it('resolveStringEvaluation returns the translated result for string flags', async () => {
@@ -97,7 +102,12 @@ it('resolveStringEvaluation returns the translated result for string flags', asy
 
   const result = await provider.resolveStringEvaluation('flag', 'red', { targetingKey: 'u' });
 
-  expect(result).toEqual({ value: 'green', variant: '0', reason: 'FALLTHROUGH' });
+  expect(result).toEqual({
+    value: 'green',
+    variant: '0',
+    reason: 'FALLTHROUGH',
+    flagMetadata: { variationIndex: 0 },
+  });
 });
 
 it('resolveNumberEvaluation returns the translated result for number flags', async () => {
@@ -111,7 +121,12 @@ it('resolveNumberEvaluation returns the translated result for number flags', asy
 
   const result = await provider.resolveNumberEvaluation('flag', 0, { targetingKey: 'u' });
 
-  expect(result).toEqual({ value: 42, variant: '2', reason: 'TARGET_MATCH' });
+  expect(result).toEqual({
+    value: 42,
+    variant: '2',
+    reason: 'TARGET_MATCH',
+    flagMetadata: { variationIndex: 2 },
+  });
 });
 
 it('resolveObjectEvaluation returns the translated result for object flags', async () => {
@@ -129,7 +144,12 @@ it('resolveObjectEvaluation returns the translated result for object flags', asy
     { targetingKey: 'u' },
   );
 
-  expect(result).toEqual({ value: { a: 1 }, variant: '3', reason: 'RULE_MATCH' });
+  expect(result).toEqual({
+    value: { a: 1 },
+    variant: '3',
+    reason: 'RULE_MATCH',
+    flagMetadata: { variationIndex: 3 },
+  });
 });
 
 it.each([
@@ -328,6 +348,11 @@ it('wraps a host logger that throws so flag evaluation does not crash', async ()
       targetingKey: 'u',
       kind: 42 as unknown as string,
     }),
-  ).resolves.toEqual({ value: true, variant: '0', reason: 'OFF' });
+  ).resolves.toEqual({
+    value: true,
+    variant: '0',
+    reason: 'OFF',
+    flagMetadata: { variationIndex: 0 },
+  });
 });
 
@@ -60,3 +60,73 @@ it('does populate the errorCode when there is an error', () => {
   });
   expect(translated.errorCode).toEqual('GENERAL');
 });
+
+it('includes the variation index in the flag metadata', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'FALLTHROUGH' },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9 });
+});
+
+it('omits the variation index from the flag metadata when there is no variation', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: null,
+      reason: { kind: 'ERROR', errorKind: 'FLAG_NOT_FOUND' },
+    }).flagMetadata,
+  ).toEqual({});
+});
+
+it('includes inExperiment in the flag metadata for experiment evaluations', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'FALLTHROUGH', inExperiment: true },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9, inExperiment: true });
+});
+
+it('omits inExperiment from the flag metadata for non-experiment evaluations', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'FALLTHROUGH', inExperiment: false },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9 });
+});
+
+it('includes the rule in the flag metadata for rule matches', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'RULE_MATCH', ruleIndex: 2, ruleId: 'the-rule-id' },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9, ruleIndex: 2, ruleId: 'the-rule-id' });
+});
+
+it('includes the prerequisite key in the flag metadata for failed prerequisites', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'PREREQUISITE_FAILED', prerequisiteKey: 'the-prerequisite-key' },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9, prerequisiteKey: 'the-prerequisite-key' });
+});
+
+it('includes the big segments status in the flag metadata', () => {
+  expect(
+    translateResult<boolean>({
+      value: true,
+      variationIndex: 9,
+      reason: { kind: 'FALLTHROUGH', bigSegmentsStatus: 'STALE' },
+    }).flagMetadata,
+  ).toEqual({ variationIndex: 9, bigSegmentsStatus: 'STALE' });
+});
@@ -1,8 +1,15 @@
 import { ErrorCode } from '@openfeature/server-sdk';
-import type { ResolutionDetails } from '@openfeature/server-sdk';
+import type { FlagMetadata, ResolutionDetails } from '@openfeature/server-sdk';
 
 import type { LDEvaluationDetail } from '@launchdarkly/js-sdk-common';
 
+const VARIATION_INDEX_KEY = 'variationIndex';
+const IN_EXPERIMENT_KEY = 'inExperiment';
+const RULE_INDEX_KEY = 'ruleIndex';
+const RULE_ID_KEY = 'ruleId';
+const PREREQUISITE_KEY_KEY = 'prerequisiteKey';
+const BIG_SEGMENTS_STATUS_KEY = 'bigSegmentsStatus';
+
 /**
  * Convert an `errorKind` into an OpenFeature `errorCode`.
  */
@@ -21,6 +28,32 @@ function translateErrorKind(errorKind: string | undefined): ErrorCode {
   }
 }
 
+/**
+ * Convert the LaunchDarkly specific parts of an evaluation into OpenFeature flag metadata.
+ */
+function translateFlagMetadata(result: LDEvaluationDetail): FlagMetadata {
+  const metadata: FlagMetadata = {};
+  if (result.variationIndex !== undefined && result.variationIndex !== null) {
+    metadata[VARIATION_INDEX_KEY] = result.variationIndex;
+  }
+  if (result.reason.inExperiment) {
+    metadata[IN_EXPERIMENT_KEY] = true;
+  }
+  if (result.reason.ruleIndex !== undefined) {
+    metadata[RULE_INDEX_KEY] = result.reason.ruleIndex;
+  }
+  if (result.reason.ruleId !== undefined) {
+    metadata[RULE_ID_KEY] = result.reason.ruleId;
+  }
+  if (result.reason.prerequisiteKey !== undefined) {
+    metadata[PREREQUISITE_KEY_KEY] = result.reason.prerequisiteKey;
+  }
+  if (result.reason.bigSegmentsStatus !== undefined) {
+    metadata[BIG_SEGMENTS_STATUS_KEY] = result.reason.bigSegmentsStatus;
+  }
+  return metadata;
+}
+
 /**
  * Translate an {@link LDEvaluationDetail} to a {@link ResolutionDetails}.
  *
@@ -30,6 +63,7 @@ export function translateResult<T>(result: LDEvaluationDetail): ResolutionDetail
     value: result.value,
     variant: result.variationIndex?.toString(),
     reason: result.reason.kind,
+    flagMetadata: translateFlagMetadata(result),
   };
 
   if (result.reason.kind === 'ERROR') {