@mastra/evals 1.2.3 → 1.2.4

package/dist/docs/references/reference-evals-context-precision.md

@@ -98,7 +98,7 @@ MAP = (1.0 + 0.67) / 2 = 0.835 ≈ **0.83**
 
 ```typescript
 const scorer = createContextPrecisionScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       // Extract context dynamically based on the query
@@ -117,7 +117,7 @@ const scorer = createContextPrecisionScorer({
 
 ```typescript
 const scorer = createContextPrecisionScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       // Simulate retrieved documents from vector database
@@ -142,7 +142,7 @@ import { createContextPrecisionScorer } from '@mastra/evals/scorers/prebuilt'
 import { myAgent } from './agent'
 
 const scorer = createContextPrecisionScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       // Extract context from agent's retrieved documents

package/dist/docs/references/reference-evals-context-relevance.md

@@ -119,7 +119,7 @@ import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 // Stricter penalty configuration
 const strictScorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'Einstein won the Nobel Prize for photoelectric effect',
@@ -137,7 +137,7 @@ const strictScorer = createContextRelevanceScorerLLM({
 
 // Lenient penalty configuration
 const lenientScorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'Einstein won the Nobel Prize for photoelectric effect',
@@ -183,7 +183,7 @@ console.log('Lenient penalties:', lenientResult.score) // Higher score, less pen
 
 ```typescript
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       // Extract context based on the query
@@ -207,7 +207,7 @@ const scorer = createContextRelevanceScorerLLM({
 
 ```typescript
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: ['Relevant information...', 'Supporting details...'],
     scale: 100, // Scale scores from 0-100 instead of 0-1
@@ -221,7 +221,7 @@ const scorer = createContextRelevanceScorerLLM({
 
 ```typescript
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       const query = input?.inputMessages?.[0]?.content || ''
@@ -248,7 +248,7 @@ This example shows excellent context relevance where all context directly suppor
 import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'Einstein won the Nobel Prize for his discovery of the photoelectric effect in 1921.',
@@ -295,7 +295,7 @@ This example shows moderate relevance with some context being irrelevant or unus
 import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'Solar eclipses occur when the Moon blocks the Sun.',
@@ -337,7 +337,7 @@ console.log(result)
 
 // With custom penalty configuration
 const customScorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'Solar eclipses occur when the Moon blocks the Sun.',
@@ -384,7 +384,7 @@ This example shows poor context relevance with mostly irrelevant information:
 import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'The Great Barrier Reef is located in Australia.',
@@ -432,7 +432,7 @@ Extract context dynamically based on the run input:
 import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       // Extract query from input
@@ -475,7 +475,7 @@ Integrate with RAG pipelines to evaluate retrieved context:
 import { createContextRelevanceScorerLLM } from '@mastra/evals'
 
 const scorer = createContextRelevanceScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     contextExtractor: (input, output) => {
       // Extract from RAG retrieval results

package/dist/docs/references/reference-evals-faithfulness.md

@@ -79,7 +79,7 @@ import { myAgent } from './agent'
 
 // Context is typically populated from agent tool calls or RAG retrieval
 const scorer = createFaithfulnessScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 const result = await runEvals({

package/dist/docs/references/reference-evals-hallucination.md

@@ -104,7 +104,7 @@ Use static context when you have known ground truth to compare against:
 import { createHallucinationScorer } from '@mastra/evals/scorers/prebuilt'
 
 const scorer = createHallucinationScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: [
       'The first iPhone was announced on January 9, 2007.',
@@ -124,7 +124,7 @@ import { createHallucinationScorer } from '@mastra/evals/scorers/prebuilt'
 import { extractToolResults } from '@mastra/evals/scorers'
 
 const scorer = createHallucinationScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     getContext: ({ run, step }) => {
       // Extract tool results as context
@@ -145,7 +145,7 @@ import { createHallucinationScorer } from '@mastra/evals/scorers/prebuilt'
 import { extractToolResults } from '@mastra/evals/scorers'
 
 const hallucinationScorer = createHallucinationScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     getContext: ({ run }) => {
       const toolResults = extractToolResults(run.output)
@@ -156,7 +156,7 @@ const hallucinationScorer = createHallucinationScorer({
 
 const agent = new Agent({
   name: 'my-agent',
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   instructions: 'You are a helpful assistant.',
   evals: {
     scorers: [hallucinationScorer],
@@ -172,7 +172,7 @@ import { createHallucinationScorer } from '@mastra/evals/scorers/prebuilt'
 import { myAgent } from './agent'
 
 const scorer = createHallucinationScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     context: ['Known fact 1', 'Known fact 2'],
   },

package/dist/docs/references/reference-evals-noise-sensitivity.md

@@ -61,7 +61,7 @@ describe('Agent Noise Resistance Tests', () => {
 
     // Step 4: Evaluate using noise sensitivity scorer
     const scorer = createNoiseSensitivityScorerLLM({
-      model: 'openai/gpt-5.4',
+      model: 'openai/gpt-5.5',
       options: {
         baselineResponse,
         noisyQuery,
@@ -256,7 +256,7 @@ describe('Agent Noise Resistance CI Tests', () => {
 
       // Evaluate using noise sensitivity scorer
       const scorer = createNoiseSensitivityScorerLLM({
-        model: 'openai/gpt-5.4',
+        model: 'openai/gpt-5.5',
         options: {
           baselineResponse: testCase.baselineResponse,
           noisyQuery: testCase.noisyQuery,
@@ -291,7 +291,7 @@ This example shows an agent that completely resists misinformation in a test sce
 import { createNoiseSensitivityScorerLLM } from '@mastra/evals'
 
 const scorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse:
       'Regular exercise improves cardiovascular health, strengthens muscles, and enhances mental wellbeing.',
@@ -337,7 +337,7 @@ This example shows an agent partially distracted by irrelevant requests:
 import { createNoiseSensitivityScorerLLM } from '@mastra/evals/scorers/prebuilt'
 
 const scorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse:
       'To bake a cake: Mix flour, sugar, eggs, and butter. Bake at 350°F for 30 minutes.',
@@ -382,7 +382,7 @@ This example shows an agent that incorporates misinformation:
 import { createNoiseSensitivityScorerLLM } from '@mastra/evals'
 
 const scorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse: 'Climate change is caused by greenhouse gas emissions from human activities.',
     noisyQuery:
@@ -428,7 +428,7 @@ import { createNoiseSensitivityScorerLLM } from '@mastra/evals'
 
 // Lenient scoring - more forgiving of minor issues
 const lenientScorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse: 'Python is a high-level programming language.',
     noisyQuery: 'What is Python? Also, snakes are dangerous!',
@@ -448,7 +448,7 @@ const lenientScorer = createNoiseSensitivityScorerLLM({
 
 // Strict scoring - harsh on any deviation
 const strictScorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse: 'Python is a high-level programming language.',
     noisyQuery: 'What is Python? Also, snakes are dangerous!',
@@ -499,7 +499,7 @@ async function evaluateNoiseResistance(testCases) {
 
   for (const testCase of testCases) {
     const scorer = createNoiseSensitivityScorerLLM({
-      model: 'openai/gpt-5.4',
+      model: 'openai/gpt-5.5',
       options: {
         baselineResponse: testCase.baseline,
         noisyQuery: testCase.noisyQuery,
@@ -546,9 +546,9 @@ import { createNoiseSensitivityScorerLLM } from '@mastra/evals'
 
 async function compareModelRobustness() {
   const models = [
-    { name: 'GPT-5.4', model: 'openai/gpt-5.4' },
+    { name: 'GPT-5.4', model: 'openai/gpt-5.5' },
     { name: 'GPT-5.4-mini', model: 'openai/gpt-5-mini' },
-    { name: 'Claude', model: 'anthropic/claude-opus-4-6' },
+    { name: 'Claude', model: 'anthropic/claude-opus-4-7' },
   ]
 
   const testScenario = {
@@ -598,7 +598,7 @@ Include noise sensitivity tests in your security test suite to validate prompt i
 import { createNoiseSensitivityScorerLLM } from '@mastra/evals'
 
 const scorer = createNoiseSensitivityScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     baselineResponse: 'I can help you with programming questions.',
     noisyQuery:

package/dist/docs/references/reference-evals-prompt-alignment.md

@@ -60,7 +60,7 @@ You can customize the Prompt Alignment Scorer by adjusting the scale parameter a
 
 ```typescript
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     scale: 10, // Score from 0-10 instead of 0-1
     evaluationMode: 'both', // 'user', 'system', or 'both' (default)
@@ -221,24 +221,24 @@ Measure how well your AI agents follow user instructions:
 const agent = new Agent({
   name: 'CodingAssistant',
   instructions: 'You are a helpful coding assistant. Always provide working code examples.',
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 // Evaluate comprehensive alignment (default)
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'both' }, // Evaluates both user intent and system guidelines
 })
 
 // Evaluate just user satisfaction
 const userScorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'user' }, // Focus only on user request fulfillment
 })
 
 // Evaluate system compliance
 const systemScorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'system' }, // Check adherence to system instructions
 })
 
@@ -290,7 +290,7 @@ for (const agent of agents) {
 import { createPromptAlignmentScorerLLM } from '@mastra/evals'
 
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 // Evaluate a code generation task
@@ -319,7 +319,7 @@ const result = await scorer.run({
 ```typescript
 // Configure scale and evaluation mode
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: {
     scale: 10, // Score from 0-10 instead of 0-1
     evaluationMode: 'both', // 'user', 'system', or 'both' (default)
@@ -328,13 +328,13 @@ const scorer = createPromptAlignmentScorerLLM({
 
 // User-only evaluation - focus on user satisfaction
 const userScorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'user' },
 })
 
 // System-only evaluation - focus on compliance
 const systemScorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'system' },
 })
 
@@ -369,7 +369,7 @@ In this example, the response fully addresses the user's prompt with all require
 import { createPromptAlignmentScorerLLM } from '@mastra/evals/scorers/prebuilt'
 
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 const inputMessages = [
@@ -417,7 +417,7 @@ In this example, the response addresses the core intent but misses some requirem
 import { createPromptAlignmentScorerLLM } from '@mastra/evals/scorers/prebuilt'
 
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 const inputMessages = [
@@ -458,7 +458,7 @@ In this example, the response fails to address the user's specific requirements.
 import { createPromptAlignmentScorerLLM } from '@mastra/evals/scorers/prebuilt'
 
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
 })
 
 const inputMessages = [
@@ -502,7 +502,7 @@ Evaluates how well the response addresses the user's request, ignoring system in
 
 ```typescript
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'user' },
 })
 
@@ -534,7 +534,7 @@ Evaluates compliance with system behavioral guidelines and constraints:
 
 ```typescript
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'system' },
 })
 
@@ -566,7 +566,7 @@ Evaluates both user intent fulfillment and system compliance with weighted scori
 
 ```typescript
 const scorer = createPromptAlignmentScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   options: { evaluationMode: 'both' }, // This is the default
 })
 

package/dist/docs/references/reference-evals-tool-call-accuracy.md

@@ -309,7 +309,7 @@ The LLM-based scorer provides:
 ```typescript
 // Basic configuration
 const basicLLMScorer = createLLMScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   availableTools: [
     { name: 'tool1', description: 'Description 1' },
     { name: 'tool2', description: 'Description 2' }
@@ -349,7 +349,7 @@ The LLM-based scorer uses AI to evaluate whether tool selections are appropriate
 
 ```typescript
 const llmScorer = createToolCallAccuracyScorerLLM({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   availableTools: [
     {
       name: 'weather-tool',
@@ -482,7 +482,7 @@ const codeScorer = createCodeScorer({
 })
 
 const llmScorer = createLLMScorer({
-  model: 'openai/gpt-5.4',
+  model: 'openai/gpt-5.5',
   availableTools: [
     { name: 'weather-tool', description: 'Get weather information' },
     { name: 'search-tool', description: 'Search the web' },

package/dist/docs/references/reference-evals-toxicity.md

@@ -86,7 +86,7 @@ import { runEvals } from '@mastra/core/evals'
 import { createToxicityScorer } from '@mastra/evals/scorers/prebuilt'
 import { myAgent } from './agent'
 
-const scorer = createToxicityScorer({ model: 'openai/gpt-5.4' })
+const scorer = createToxicityScorer({ model: 'openai/gpt-5.5' })
 
 const result = await runEvals({
   data: [