@mastra/evals 1.5.0 → 1.5.1-alpha.0

package/dist/scorers/prebuilt/index.cjs

@@ -1,7 +1,7 @@
 'use strict';
 
-var chunkNXVFY4CK_cjs = require('../../chunk-NXVFY4CK.cjs');
-var chunkUNQXHPOD_cjs = require('../../chunk-UNQXHPOD.cjs');
+var chunkGGHVFNVI_cjs = require('../../chunk-GGHVFNVI.cjs');
+var chunkIZLA36WC_cjs = require('../../chunk-IZLA36WC.cjs');
 var evals = require('@mastra/core/evals');
 var nlp = require('compromise');
 var keyword_extractor = require('keyword-extractor');
@@ -251,7 +251,7 @@ function createAnswerRelevancyScorer({
     description: "Extract relevant statements from the LLM output",
     outputSchema: extractOutputSchema,
     createPrompt: ({ run }) => {
-      const assistantMessage = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const assistantMessage = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       return createExtractPrompt(assistantMessage);
     }
   }).analyze({
@@ -284,7 +284,7 @@ function createAnswerRelevancyScorer({
       ]
     },
     createPrompt: ({ run, results }) => {
-      const input = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const input = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
       return createScorePrompt(JSON.stringify(input), results.preprocessStepResult?.statements || []);
     }
   }).generateScore(({ results }) => {
@@ -301,13 +301,13 @@ function createAnswerRelevancyScorer({
       }
     }
     const score = relevancyCount / numberOfResults;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score * options.scale);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score * options.scale);
   }).generateReason({
     description: "Reason about the results",
     createPrompt: ({ run, results, score }) => {
       return createReasonPrompt({
-        input: chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "",
-        output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
+        input: chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "",
+        output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
         score,
         results: results.analyzeStepResult.results,
         scale: options.scale
@@ -582,7 +582,7 @@ function createAnswerSimilarityScorer({
           groundTruth: ""
         });
       }
-      const output = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const output = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const groundTruth = typeof run.groundTruth === "string" ? run.groundTruth : JSON.stringify(run.groundTruth);
       return createExtractPrompt2({
         output,
@@ -640,14 +640,14 @@ function createAnswerSimilarityScorer({
     );
     score -= extraInfoPenalty;
     score = Math.max(0, Math.min(1, score));
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score * mergedOptions.scale);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score * mergedOptions.scale);
   }).generateReason({
     description: "Generate explanation of similarity score",
     createPrompt: ({ run, results, score }) => {
       if (!run.groundTruth) {
         return "No ground truth was provided for comparison. Score is 0 by default.";
       }
-      const output = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const output = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const groundTruth = typeof run.groundTruth === "string" ? run.groundTruth : JSON.stringify(run.groundTruth);
       return createReasonPrompt2({
         output,
@@ -849,7 +849,7 @@ function createFaithfulnessScorer({
       ]
     },
     createPrompt: ({ run }) => {
-      const prompt = createFaithfulnessExtractPrompt({ output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" });
+      const prompt = createFaithfulnessExtractPrompt({ output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" });
       return prompt;
     }
   }).analyze({
@@ -896,13 +896,13 @@ function createFaithfulnessScorer({
       return 0;
     }
     const score = supportedClaims / totalClaims * (options?.scale || 1);
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score);
   }).generateReason({
     description: "Reason about the results",
     createPrompt: ({ run, results, score }) => {
       const prompt = createFaithfulnessReasonPrompt({
-        input: chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "",
-        output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
+        input: chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "",
+        output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
         context: options?.context ?? getToolInvocationContext(run.output),
         score,
         scale: options?.scale || 1,
@@ -1047,7 +1047,7 @@ function createBiasScorer({ model, options }) {
         "opinions"
       ]
     },
-    createPrompt: ({ run }) => createBiasExtractPrompt({ output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" })
+    createPrompt: ({ run }) => createBiasExtractPrompt({ output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" })
   }).analyze({
     description: "Score the relevance of the statements to the input",
     outputSchema: {
@@ -1079,7 +1079,7 @@ function createBiasScorer({ model, options }) {
     },
     createPrompt: ({ run, results }) => {
       const prompt = createBiasAnalyzePrompt({
-        output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
+        output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
         opinions: results.preprocessStepResult?.opinions || []
       });
       return prompt;
@@ -1090,7 +1090,7 @@ function createBiasScorer({ model, options }) {
     }
     const biasedVerdicts = results.analyzeStepResult.results.filter((v) => v.result.toLowerCase() === "yes");
     const score = biasedVerdicts.length / results.analyzeStepResult.results.length;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score * (options?.scale || 1));
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score * (options?.scale || 1));
   }).generateReason({
     description: "Reason about the results",
     createPrompt: ({ score, results }) => {
@@ -1321,7 +1321,7 @@ function createHallucinationScorer({
       ]
     },
     createPrompt: ({ run }) => {
-      const prompt = createHallucinationExtractPrompt({ output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" });
+      const prompt = createHallucinationExtractPrompt({ output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "" });
       return prompt;
     }
   }).analyze({
@@ -1377,7 +1377,7 @@ function createHallucinationScorer({
       return 0;
     }
     const score = contradictedStatements / totalStatements * (options?.scale || 1);
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score);
   }).generateReason({
     description: "Reason about the results",
     createPrompt: async ({ run, results, score }) => {
@@ -1388,8 +1388,8 @@ function createHallucinationScorer({
         context = options?.context ?? [];
       }
       const prompt = createHallucinationReasonPrompt({
-        input: chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "",
-        output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
+        input: chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "",
+        output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "",
         context,
         score,
         scale: options?.scale || 1,
@@ -1529,8 +1529,8 @@ function createToxicityScorer({
     },
     createPrompt: ({ run }) => {
       const prompt = createToxicityAnalyzePrompt({
-        input: chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "",
-        output: chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? ""
+        input: chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "",
+        output: chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? ""
       });
       return prompt;
     }
@@ -1546,7 +1546,7 @@ function createToxicityScorer({
       }
     }
     const score = toxicityCount / numberOfVerdicts;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score * (options?.scale || 1));
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score * (options?.scale || 1));
   }).generateReason({
     description: "Reason about the results",
     createPrompt: ({ results, score }) => {
@@ -1707,7 +1707,7 @@ function createToolCallAccuracyScorerLLM({ model, availableTools }) {
     if (isInputInvalid || isOutputInvalid) {
       throw new Error("Input and output messages cannot be null or empty");
     }
-    const { tools: actualTools, toolCallInfos } = chunkUNQXHPOD_cjs.extractToolCalls(run.output);
+    const { tools: actualTools, toolCallInfos } = chunkIZLA36WC_cjs.extractToolCalls(run.output);
     return {
       actualTools,
       hasToolCalls: actualTools.length > 0,
@@ -1717,8 +1717,8 @@ function createToolCallAccuracyScorerLLM({ model, availableTools }) {
     description: "Analyze the appropriateness of tool selections",
     outputSchema: analyzeOutputSchema2,
     createPrompt: ({ run, results }) => {
-      const userInput = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const agentResponse = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const userInput = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const agentResponse = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const toolsCalled = results.preprocessStepResult?.actualTools || [];
       return createAnalyzePrompt2({
         userInput,
@@ -1735,11 +1735,11 @@ function createToolCallAccuracyScorerLLM({ model, availableTools }) {
     }
     const appropriateToolCalls = evaluations.filter((e) => e.wasAppropriate).length;
     const totalToolCalls = evaluations.length;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(appropriateToolCalls / totalToolCalls);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(appropriateToolCalls / totalToolCalls);
   }).generateReason({
     description: "Generate human-readable explanation of tool selection evaluation",
     createPrompt: ({ run, results, score }) => {
-      const userInput = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const userInput = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
       const evaluations = results.analyzeStepResult?.evaluations || [];
       const missingTools = results.analyzeStepResult?.missingTools || [];
       return createReasonPrompt3({
@@ -1969,7 +1969,7 @@ var getContext = ({
   output,
   options
 }) => {
-  if (options.contextExtractor && chunkUNQXHPOD_cjs.isScorerRunInputForAgent(input) && chunkUNQXHPOD_cjs.isScorerRunOutputForAgent(output)) {
+  if (options.contextExtractor && chunkIZLA36WC_cjs.isScorerRunInputForAgent(input) && chunkIZLA36WC_cjs.isScorerRunOutputForAgent(output)) {
     return options.contextExtractor(input, output);
   }
   return options.context ?? [];
@@ -1997,8 +1997,8 @@ function createContextRelevanceScorerLLM({
     description: "Analyze the relevance and utility of provided context",
     outputSchema: analyzeOutputSchema3,
     createPrompt: ({ run }) => {
-      const userQuery = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const agentResponse = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const userQuery = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const agentResponse = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const context = getContext({ input: run.input, output: run.output, options });
       if (context.length === 0) {
         return createAnalyzePrompt3({
@@ -2046,11 +2046,11 @@ function createContextRelevanceScorerLLM({
     const missingContextPenalty = Math.min(missingContext.length * missingPenaltyRate, maxMissingPenalty);
     const finalScore = Math.max(0, relevanceScore - usagePenalty - missingContextPenalty);
     const scaledScore = finalScore * (options.scale || 1);
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(scaledScore);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(scaledScore);
   }).generateReason({
     description: "Generate human-readable explanation of context relevance evaluation",
     createPrompt: ({ run, results, score }) => {
-      const userQuery = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const userQuery = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
       const context = getContext({ input: run.input, output: run.output, options });
       if (context.length === 0) {
         return `No context was available for evaluation. The agent response was generated without any supporting context. Score: ${score}`;
@@ -2225,7 +2225,7 @@ var getContext2 = ({
   output,
   options
 }) => {
-  if (options.contextExtractor && chunkUNQXHPOD_cjs.isScorerRunInputForAgent(input) && chunkUNQXHPOD_cjs.isScorerRunOutputForAgent(output)) {
+  if (options.contextExtractor && chunkIZLA36WC_cjs.isScorerRunInputForAgent(input) && chunkIZLA36WC_cjs.isScorerRunOutputForAgent(output)) {
     return options.contextExtractor(input, output);
   }
   return options.context ?? [];
@@ -2253,8 +2253,8 @@ function createContextPrecisionScorer({
     description: "Evaluate the relevance of each context piece for generating the expected output",
     outputSchema: contextRelevanceOutputSchema,
     createPrompt: ({ run }) => {
-      const input = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const output = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const input = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const output = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const context = getContext2({ input: run.input, output: run.output, options });
       if (context.length === 0) {
         throw new Error("No context available for evaluation");
@@ -2287,12 +2287,12 @@ function createContextPrecisionScorer({
     }
     const map = sumPrecision / relevantCount;
     const score = map * (options.scale || 1);
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(score);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(score);
   }).generateReason({
     description: "Reason about the context precision results",
     createPrompt: ({ run, results, score }) => {
-      const input = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const output = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const input = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const output = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       const context = getContext2({ input: run.input, output: run.output, options });
       return createContextPrecisionReasonPrompt({
         input,
@@ -2590,8 +2590,8 @@ function createNoiseSensitivityScorerLLM({
     description: "Analyze the impact of noise on agent response quality",
     outputSchema: analyzeOutputSchema4,
     createPrompt: ({ run }) => {
-      const originalQuery = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const noisyResponse = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const originalQuery = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const noisyResponse = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       if (!originalQuery || !noisyResponse) {
         throw new Error("Both original query and noisy response are required for evaluation");
       }
@@ -2634,11 +2634,11 @@ function createNoiseSensitivityScorerLLM({
     const majorIssues = analysisResult.majorIssues || [];
     const issuesPenalty = Math.min(majorIssues.length * majorIssuePenaltyRate, maxMajorIssuePenalty);
     finalScore = Math.max(0, finalScore - issuesPenalty);
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(finalScore);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(finalScore);
   }).generateReason({
     description: "Generate human-readable explanation of noise sensitivity evaluation",
     createPrompt: ({ run, results, score }) => {
-      const originalQuery = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const originalQuery = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
       const analysisResult = results.analyzeStepResult;
       if (!analysisResult) {
         throw new Error("Analysis step failed to produce results for reason generation");
@@ -3050,9 +3050,9 @@ function createPromptAlignmentScorerLLM({
     description: "Analyze prompt-response alignment across multiple dimensions",
     outputSchema: analyzeOutputSchema5,
     createPrompt: ({ run }) => {
-      const userPrompt = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const systemPrompt = chunkUNQXHPOD_cjs.getCombinedSystemPrompt(run.input) ?? "";
-      const agentResponse = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
+      const userPrompt = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const systemPrompt = chunkIZLA36WC_cjs.getCombinedSystemPrompt(run.input) ?? "";
+      const agentResponse = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output) ?? "";
       if (evaluationMode === "user" && !userPrompt) {
         throw new Error("User prompt is required for user prompt alignment scoring");
       }
@@ -3088,12 +3088,12 @@ function createPromptAlignmentScorerLLM({
       weightedScore = userScore * SCORING_WEIGHTS.BOTH.USER_WEIGHT + systemScore * SCORING_WEIGHTS.BOTH.SYSTEM_WEIGHT;
     }
     const finalScore = weightedScore * scale;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(finalScore);
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(finalScore);
   }).generateReason({
     description: "Generate human-readable explanation of prompt alignment evaluation",
     createPrompt: ({ run, results, score }) => {
-      const userPrompt = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const systemPrompt = chunkUNQXHPOD_cjs.getCombinedSystemPrompt(run.input) ?? "";
+      const userPrompt = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const systemPrompt = chunkIZLA36WC_cjs.getCombinedSystemPrompt(run.input) ?? "";
       const analysis = results.analyzeStepResult;
       if (!analysis) {
         return `Unable to analyze prompt alignment. Score: ${score}`;
@@ -3247,7 +3247,7 @@ function toCriterionInputs(criteria) {
   return criteria.map((c) => ({ criterion: c.description, required: c.required !== false }));
 }
 function getOutputText(run) {
-  const fromOutput = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output);
+  const fromOutput = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output);
   if (fromOutput) return fromOutput;
   if (run.input && typeof run.input === "object" && typeof run.input.currentText === "string") {
     return run.input.currentText;
@@ -3258,7 +3258,7 @@ function getTaskText(run) {
   if (run.input && typeof run.input === "object" && typeof run.input.originalTask === "string") {
     return run.input.originalTask;
   }
-  return chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+  return chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
 }
 function createRubricScorer({
   model,
@@ -3540,8 +3540,8 @@ function createTrajectoryAccuracyScorerLLM({
     description: "Analyze the quality and appropriateness of the agent trajectory",
     outputSchema: analyzeOutputSchema7,
     createPrompt: ({ run, results }) => {
-      const userInput = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
-      const agentResponse = chunkUNQXHPOD_cjs.getAssistantMessageFromRunOutput(run.output.rawOutput) ?? "";
+      const userInput = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const agentResponse = chunkIZLA36WC_cjs.getAssistantMessageFromRunOutput(run.output.rawOutput) ?? "";
       return createAnalyzePrompt7({
         userInput,
         agentResponse,
@@ -3566,11 +3566,11 @@ function createTrajectoryAccuracyScorerLLM({
     const necessityScore = necessarySteps / totalSteps;
     const orderScore = orderedSteps / totalSteps;
     const score = necessityScore * 0.6 + orderScore * 0.3 - missingPenalty * 0.1;
-    return chunkUNQXHPOD_cjs.roundToTwoDecimals(Math.max(0, Math.min(1, score)));
+    return chunkIZLA36WC_cjs.roundToTwoDecimals(Math.max(0, Math.min(1, score)));
   }).generateReason({
     description: "Generate human-readable explanation of trajectory evaluation",
     createPrompt: ({ run, results, score }) => {
-      const userInput = chunkUNQXHPOD_cjs.getUserMessageFromRunInput(run.input) ?? "";
+      const userInput = chunkIZLA36WC_cjs.getUserMessageFromRunInput(run.input) ?? "";
       const stepEvaluations = results.analyzeStepResult?.stepEvaluations || [];
       const missingSteps = results.analyzeStepResult?.missingSteps || [];
       const extraSteps = results.analyzeStepResult?.extraSteps || [];
@@ -3633,18 +3633,18 @@ function createCompletenessScorer() {
     type: "agent"
   }).preprocess(async ({ run }) => {
     const isInputInvalid = !run.input || run.input.inputMessages.some((i) => {
-      const content = chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i);
+      const content = chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i);
       return content === null || content === void 0;
     });
     const isOutputInvalid = !run.output || run.output.some((i) => {
-      const content = chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i);
+      const content = chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i);
       return content === null || content === void 0;
     });
     if (isInputInvalid || isOutputInvalid) {
       throw new Error("Inputs cannot be null or undefined");
     }
-    const input = run.input?.inputMessages.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
-    const output = run.output?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const input = run.input?.inputMessages.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const output = run.output?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
     const inputToProcess = input;
     const outputToProcess = output;
     const inputDoc = nlp__default.default(inputToProcess.trim());
@@ -3749,8 +3749,8 @@ function createTextualDifferenceScorer() {
     description: "Calculate textual difference between input and output using sequence matching algorithms.",
     type: "agent"
   }).preprocess(async ({ run }) => {
-    const input = run.input?.inputMessages?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
-    const output = run.output?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const input = run.input?.inputMessages?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const output = run.output?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
     const ratio = calculateRatio(input, output);
     const changes = countChanges(input, output);
     const maxLength = Math.max(input.length, output.length);
@@ -3773,8 +3773,8 @@ function createKeywordCoverageScorer() {
     description: 'Leverage the nlp method from "compromise" to extract elements from the input and output and calculate the coverage.',
     type: "agent"
   }).preprocess(async ({ run }) => {
-    const input = run.input?.inputMessages?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
-    const output = run.output?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const input = run.input?.inputMessages?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const output = run.output?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
     if (!input && !output) {
       return {
         result: {
@@ -3827,8 +3827,8 @@ function createContentSimilarityScorer({ ignoreCase, ignoreWhitespace } = { igno
     description: "Calculates content similarity between input and output messages using string comparison algorithms.",
     type: "agent"
   }).preprocess(async ({ run }) => {
-    let processedInput = run.input?.inputMessages.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
-    let processedOutput = run.output.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    let processedInput = run.input?.inputMessages.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    let processedOutput = run.output.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
     if (ignoreCase) {
       processedInput = processedInput.toLowerCase();
       processedOutput = processedOutput.toLowerCase();
@@ -3858,7 +3858,7 @@ function createToneScorer(config = {}) {
     type: "agent"
   }).preprocess(async ({ run }) => {
     const sentiment = new Sentiment__default.default();
-    const agentMessage = run.output?.map((i) => chunkUNQXHPOD_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
+    const agentMessage = run.output?.map((i) => chunkIZLA36WC_cjs.getTextContentFromMastraDBMessage(i)).join(", ") || "";
     const responseSentiment = sentiment.analyze(agentMessage);
     if (referenceTone) {
       const referenceSentiment = sentiment.analyze(referenceTone);
@@ -3945,7 +3945,7 @@ function createToolCallAccuracyScorerCode(options) {
     if (isInputInvalid || isOutputInvalid) {
       throw new Error("Input and output messages cannot be null or empty");
     }
-    const { tools: actualTools, toolCallInfos } = chunkUNQXHPOD_cjs.extractToolCalls(run.output);
+    const { tools: actualTools, toolCallInfos } = chunkIZLA36WC_cjs.extractToolCalls(run.output);
     const correctToolCalled = expectedTool ? strictMode ? actualTools.length === 1 && actualTools[0] === expectedTool : actualTools.includes(expectedTool) : false;
     return {
       expectedTool,
@@ -4020,7 +4020,7 @@ function createTrajectoryAccuracyScorerCode(options = {}) {
     const itemExpectation = run.expectedTrajectory;
     const effectiveOrdering = itemExpectation?.ordering ?? ordering;
     const effectiveAllowRepeated = itemExpectation?.allowRepeatedSteps ?? allowRepeatedSteps;
-    const comparison = chunkUNQXHPOD_cjs.compareTrajectories(
+    const comparison = chunkIZLA36WC_cjs.compareTrajectories(
       actualTrajectory,
       { steps: resolvedExpectedSteps },
       {
@@ -4078,7 +4078,7 @@ function evaluateNestedExpectations(expectedSteps, actualSteps, weights = { accu
     const childConfig = expectedStep.children;
     let accuracy;
     if (childConfig.steps && childConfig.steps.length > 0) {
-      accuracy = chunkUNQXHPOD_cjs.compareTrajectories(
+      accuracy = chunkIZLA36WC_cjs.compareTrajectories(
         childTrajectory,
         { steps: childConfig.steps },
         {
@@ -4088,18 +4088,18 @@ function evaluateNestedExpectations(expectedSteps, actualSteps, weights = { accu
       );
     }
     const hasEfficiencyConfig = childConfig.maxSteps !== void 0 || childConfig.maxTotalTokens !== void 0 || childConfig.maxTotalDurationMs !== void 0 || childConfig.noRedundantCalls !== void 0;
-    const efficiency = hasEfficiencyConfig ? chunkUNQXHPOD_cjs.checkTrajectoryEfficiency(childTrajectory, {
+    const efficiency = hasEfficiencyConfig ? chunkIZLA36WC_cjs.checkTrajectoryEfficiency(childTrajectory, {
       maxSteps: childConfig.maxSteps,
       maxTotalTokens: childConfig.maxTotalTokens,
       maxTotalDurationMs: childConfig.maxTotalDurationMs,
       noRedundantCalls: childConfig.noRedundantCalls ?? true
     }) : void 0;
     const hasBlacklistConfig = childConfig.blacklistedTools && childConfig.blacklistedTools.length > 0 || childConfig.blacklistedSequences && childConfig.blacklistedSequences.length > 0;
-    const blacklist = hasBlacklistConfig ? chunkUNQXHPOD_cjs.checkTrajectoryBlacklist(childTrajectory, {
+    const blacklist = hasBlacklistConfig ? chunkIZLA36WC_cjs.checkTrajectoryBlacklist(childTrajectory, {
       blacklistedTools: childConfig.blacklistedTools,
       blacklistedSequences: childConfig.blacklistedSequences
     }) : void 0;
-    const toolFailures = chunkUNQXHPOD_cjs.analyzeToolFailures(childTrajectory, {
+    const toolFailures = chunkIZLA36WC_cjs.analyzeToolFailures(childTrajectory, {
       maxRetriesPerTool: childConfig.maxRetriesPerTool ?? 2
     });
     const nested = childConfig.steps ? evaluateNestedExpectations(childConfig.steps, actualStep.children, weights) : [];
@@ -4164,7 +4164,7 @@ function createTrajectoryScorerCode(options = {}) {
     }
     let accuracy;
     if (config.steps && config.steps.length > 0) {
-      accuracy = chunkUNQXHPOD_cjs.compareTrajectories(
+      accuracy = chunkIZLA36WC_cjs.compareTrajectories(
         actualTrajectory,
         { steps: config.steps },
         {
@@ -4174,18 +4174,18 @@ function createTrajectoryScorerCode(options = {}) {
       );
     }
     const hasEfficiencyConfig = config.maxSteps !== void 0 || config.maxTotalTokens !== void 0 || config.maxTotalDurationMs !== void 0 || config.noRedundantCalls !== void 0;
-    const efficiency = hasEfficiencyConfig ? chunkUNQXHPOD_cjs.checkTrajectoryEfficiency(actualTrajectory, {
+    const efficiency = hasEfficiencyConfig ? chunkIZLA36WC_cjs.checkTrajectoryEfficiency(actualTrajectory, {
       maxSteps: config.maxSteps,
       maxTotalTokens: config.maxTotalTokens,
       maxTotalDurationMs: config.maxTotalDurationMs,
       noRedundantCalls: config.noRedundantCalls ?? true
     }) : void 0;
     const hasBlacklistConfig = config.blacklistedTools && config.blacklistedTools.length > 0 || config.blacklistedSequences && config.blacklistedSequences.length > 0;
-    const blacklist = hasBlacklistConfig ? chunkUNQXHPOD_cjs.checkTrajectoryBlacklist(actualTrajectory, {
+    const blacklist = hasBlacklistConfig ? chunkIZLA36WC_cjs.checkTrajectoryBlacklist(actualTrajectory, {
       blacklistedTools: config.blacklistedTools,
       blacklistedSequences: config.blacklistedSequences
     }) : void 0;
-    const toolFailures = chunkUNQXHPOD_cjs.analyzeToolFailures(actualTrajectory, {
+    const toolFailures = chunkIZLA36WC_cjs.analyzeToolFailures(actualTrajectory, {
       maxRetriesPerTool: config.maxRetriesPerTool ?? 2
     });
     const nested = config.steps && config.steps.length > 0 ? evaluateNestedExpectations(config.steps, actualTrajectory.steps, w) : void 0;
@@ -4305,51 +4305,51 @@ function createTrajectoryScorerCode(options = {}) {
 
 Object.defineProperty(exports, "calledTool", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.calledTool; }
+  get: function () { return chunkGGHVFNVI_cjs.calledTool; }
 });
 Object.defineProperty(exports, "checks", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.checks; }
+  get: function () { return chunkGGHVFNVI_cjs.checks; }
 });
 Object.defineProperty(exports, "didNotCall", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.didNotCall; }
+  get: function () { return chunkGGHVFNVI_cjs.didNotCall; }
 });
 Object.defineProperty(exports, "equals", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.equals; }
+  get: function () { return chunkGGHVFNVI_cjs.equals; }
 });
 Object.defineProperty(exports, "excludes", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.excludes; }
+  get: function () { return chunkGGHVFNVI_cjs.excludes; }
 });
 Object.defineProperty(exports, "includes", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.includes; }
+  get: function () { return chunkGGHVFNVI_cjs.includes; }
 });
 Object.defineProperty(exports, "matches", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.matches; }
+  get: function () { return chunkGGHVFNVI_cjs.matches; }
 });
 Object.defineProperty(exports, "maxToolCalls", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.maxToolCalls; }
+  get: function () { return chunkGGHVFNVI_cjs.maxToolCalls; }
 });
 Object.defineProperty(exports, "noToolErrors", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.noToolErrors; }
+  get: function () { return chunkGGHVFNVI_cjs.noToolErrors; }
 });
 Object.defineProperty(exports, "similarity", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.similarity; }
+  get: function () { return chunkGGHVFNVI_cjs.similarity; }
 });
 Object.defineProperty(exports, "toolOrder", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.toolOrder; }
+  get: function () { return chunkGGHVFNVI_cjs.toolOrder; }
 });
 Object.defineProperty(exports, "usedNoTools", {
   enumerable: true,
-  get: function () { return chunkNXVFY4CK_cjs.usedNoTools; }
+  get: function () { return chunkGGHVFNVI_cjs.usedNoTools; }
 });
 exports.ANSWER_RELEVANCY_AGENT_INSTRUCTIONS = ANSWER_RELEVANCY_AGENT_INSTRUCTIONS;
 exports.ANSWER_SIMILARITY_DEFAULT_OPTIONS = ANSWER_SIMILARITY_DEFAULT_OPTIONS;

package/dist/scorers/prebuilt/index.js

@@ -1,5 +1,5 @@
-export { calledTool, checks, didNotCall, equals, excludes, includes, matches, maxToolCalls, noToolErrors, similarity, toolOrder, usedNoTools } from '../../chunk-QRAONHNV.js';
-import { getAssistantMessageFromRunOutput, getUserMessageFromRunInput, roundToTwoDecimals, extractToolCalls, getCombinedSystemPrompt, getTextContentFromMastraDBMessage, compareTrajectories, checkTrajectoryEfficiency, checkTrajectoryBlacklist, analyzeToolFailures, isScorerRunInputForAgent, isScorerRunOutputForAgent } from '../../chunk-BE5F2OUQ.js';
+export { calledTool, checks, didNotCall, equals, excludes, includes, matches, maxToolCalls, noToolErrors, similarity, toolOrder, usedNoTools } from '../../chunk-WEADJCUA.js';
+import { getAssistantMessageFromRunOutput, getUserMessageFromRunInput, roundToTwoDecimals, extractToolCalls, getCombinedSystemPrompt, getTextContentFromMastraDBMessage, compareTrajectories, checkTrajectoryEfficiency, checkTrajectoryBlacklist, analyzeToolFailures, isScorerRunInputForAgent, isScorerRunOutputForAgent } from '../../chunk-UJ4WCQ3F.js';
 import { createScorer } from '@mastra/core/evals';
 import nlp from 'compromise';
 import keyword_extractor from 'keyword-extractor';