@wix/evalforge-types 0.4.0 → 0.6.0

package/build/index.mjs

@@ -516,128 +516,46 @@ var LLMTraceSchema = z21.object({
 });
 
 // src/evaluation/eval-result.ts
-import { z as z22 } from "zod";
-var AssertionResultStatus = /* @__PURE__ */ ((AssertionResultStatus2) => {
-  AssertionResultStatus2["PASSED"] = "passed";
-  AssertionResultStatus2["FAILED"] = "failed";
-  AssertionResultStatus2["SKIPPED"] = "skipped";
-  AssertionResultStatus2["ERROR"] = "error";
-  return AssertionResultStatus2;
-})(AssertionResultStatus || {});
-var AssertionResultSchema = z22.object({
-  id: z22.string(),
-  assertionId: z22.string(),
-  assertionType: z22.string(),
-  assertionName: z22.string(),
-  status: z22.enum(AssertionResultStatus),
-  message: z22.string().optional(),
-  expected: z22.string().optional(),
-  actual: z22.string().optional(),
-  duration: z22.number().optional(),
-  details: z22.record(z22.string(), z22.unknown()).optional(),
-  llmTraceSteps: z22.array(LLMTraceStepSchema).optional()
-});
-var EvalRunResultSchema = z22.object({
-  id: z22.string(),
-  targetId: z22.string(),
-  targetName: z22.string().optional(),
-  scenarioId: z22.string(),
-  scenarioName: z22.string(),
-  modelConfig: ModelConfigSchema.optional(),
-  assertionResults: z22.array(AssertionResultSchema),
-  metrics: EvalMetricsSchema.optional(),
-  passed: z22.number(),
-  failed: z22.number(),
-  passRate: z22.number(),
-  duration: z22.number(),
-  outputText: z22.string().optional(),
-  files: z22.array(ExpectedFileSchema).optional(),
-  startedAt: z22.string().optional(),
-  completedAt: z22.string().optional(),
-  llmTrace: LLMTraceSchema.optional()
-});
-var PromptResultSchema = z22.object({
-  text: z22.string(),
-  files: z22.array(z22.unknown()).optional(),
-  finishReason: z22.string().optional(),
-  reasoning: z22.string().optional(),
-  reasoningDetails: z22.unknown().optional(),
-  toolCalls: z22.array(z22.unknown()).optional(),
-  toolResults: z22.array(z22.unknown()).optional(),
-  warnings: z22.array(z22.unknown()).optional(),
-  sources: z22.array(z22.unknown()).optional(),
-  steps: z22.array(z22.unknown()),
-  generationTimeMs: z22.number(),
-  prompt: z22.string(),
-  systemPrompt: z22.string(),
-  usage: z22.object({
-    totalTokens: z22.number().optional(),
-    totalMicrocentsSpent: z22.number().optional()
-  })
-});
-var EvaluationResultSchema = z22.object({
-  id: z22.string(),
-  runId: z22.string(),
-  timestamp: z22.number(),
-  promptResult: PromptResultSchema,
-  testResults: z22.array(z22.unknown()),
-  tags: z22.array(z22.string()).optional(),
-  feedback: z22.string().optional(),
-  score: z22.number(),
-  suiteId: z22.string().optional()
-});
-var LeanEvaluationResultSchema = z22.object({
-  id: z22.string(),
-  runId: z22.string(),
-  timestamp: z22.number(),
-  tags: z22.array(z22.string()).optional(),
-  scenarioId: z22.string(),
-  scenarioVersion: z22.number().optional(),
-  targetId: z22.string(),
-  targetVersion: z22.number().optional(),
-  suiteId: z22.string().optional(),
-  score: z22.number(),
-  time: z22.number().optional(),
-  microcentsSpent: z22.number().optional()
-});
+import { z as z24 } from "zod";
 
 // src/evaluation/eval-run.ts
-import { z as z24 } from "zod";
+import { z as z23 } from "zod";
 
 // src/evaluation/live-trace.ts
-import { z as z23 } from "zod";
+import { z as z22 } from "zod";
 var LiveTraceEventType = /* @__PURE__ */ ((LiveTraceEventType2) => {
   LiveTraceEventType2["THINKING"] = "thinking";
   LiveTraceEventType2["TOOL_USE"] = "tool_use";
   LiveTraceEventType2["COMPLETION"] = "completion";
   LiveTraceEventType2["TOOL_RESULT"] = "tool_result";
+  LiveTraceEventType2["DIAGNOSTIC"] = "diagnostic";
   return LiveTraceEventType2;
 })(LiveTraceEventType || {});
-var LiveTraceEventSchema = z23.object({
+var LiveTraceEventSchema = z22.object({
   /** The evaluation run ID */
-  evalRunId: z23.string(),
+  evalRunId: z22.string(),
   /** The scenario ID being executed */
-  scenarioId: z23.string(),
+  scenarioId: z22.string(),
   /** The scenario name for display */
-  scenarioName: z23.string(),
+  scenarioName: z22.string(),
   /** The target ID (skill, agent, etc.) */
-  targetId: z23.string(),
+  targetId: z22.string(),
   /** The target name for display */
-  targetName: z23.string(),
+  targetName: z22.string(),
   /** Step number in the current scenario execution */
-  stepNumber: z23.number(),
+  stepNumber: z22.number(),
   /** Type of trace event */
-  type: z23.enum(LiveTraceEventType),
+  type: z22.enum(LiveTraceEventType),
   /** Tool name if this is a tool_use event */
-  toolName: z23.string().optional(),
+  toolName: z22.string().optional(),
   /** Tool arguments preview (truncated JSON) */
-  toolArgs: z23.string().optional(),
+  toolArgs: z22.string().optional(),
   /** Output preview (truncated text) */
-  outputPreview: z23.string().optional(),
+  outputPreview: z22.string().optional(),
   /** Timestamp when this event occurred */
-  timestamp: z23.string(),
+  timestamp: z22.string(),
   /** Whether this is the final event for this scenario */
-  isComplete: z23.boolean()
+  isComplete: z22.boolean()
 });
 var TRACE_EVENT_PREFIX = "TRACE_EVENT:";
 function parseTraceEventLine(line) {
@@ -665,14 +583,14 @@ var TriggerType = /* @__PURE__ */ ((TriggerType2) => {
   TriggerType2["MANUAL"] = "MANUAL";
   return TriggerType2;
 })(TriggerType || {});
-var TriggerMetadataSchema = z24.object({
-  version: z24.string().optional(),
-  resourceUpdated: z24.array(z24.string()).optional()
+var TriggerMetadataSchema = z23.object({
+  version: z23.string().optional(),
+  resourceUpdated: z23.array(z23.string()).optional()
 });
-var TriggerSchema = z24.object({
-  id: z24.string(),
+var TriggerSchema = z23.object({
+  id: z23.string(),
   metadata: TriggerMetadataSchema.optional(),
-  type: z24.enum(TriggerType)
+  type: z23.enum(TriggerType)
 });
 var FailureCategory = /* @__PURE__ */ ((FailureCategory2) => {
   FailureCategory2["MISSING_FILE"] = "missing_file";
@@ -690,89 +608,89 @@ var FailureSeverity = /* @__PURE__ */ ((FailureSeverity2) => {
   FailureSeverity2["LOW"] = "low";
   return FailureSeverity2;
 })(FailureSeverity || {});
-var DiffLineTypeSchema = z24.enum(["added", "removed", "unchanged"]);
-var DiffLineSchema = z24.object({
+var DiffLineTypeSchema = z23.enum(["added", "removed", "unchanged"]);
+var DiffLineSchema = z23.object({
   type: DiffLineTypeSchema,
-  content: z24.string(),
-  lineNumber: z24.number()
-});
-var DiffContentSchema = z24.object({
-  path: z24.string(),
-  expected: z24.string(),
-  actual: z24.string(),
-  diffLines: z24.array(DiffLineSchema)
-});
-var CommandExecutionSchema = z24.object({
-  command: z24.string(),
-  exitCode: z24.number(),
-  output: z24.string().optional(),
-  duration: z24.number()
-});
-var FileModificationSchema = z24.object({
-  path: z24.string(),
-  action: z24.enum(["created", "modified", "deleted"])
-});
-var ApiCallSchema = z24.object({
-  endpoint: z24.string(),
-  tokensUsed: z24.number(),
-  duration: z24.number()
-});
-var ExecutionTraceSchema = z24.object({
-  commands: z24.array(CommandExecutionSchema),
-  filesModified: z24.array(FileModificationSchema),
-  apiCalls: z24.array(ApiCallSchema),
-  totalDuration: z24.number()
-});
-var FailureAnalysisSchema = z24.object({
-  category: z24.enum(FailureCategory),
-  severity: z24.enum(FailureSeverity),
-  summary: z24.string(),
-  details: z24.string(),
-  rootCause: z24.string(),
-  suggestedFix: z24.string(),
-  relatedAssertions: z24.array(z24.string()),
-  codeSnippet: z24.string().optional(),
-  similarIssues: z24.array(z24.string()).optional(),
-  patternId: z24.string().optional(),
+  content: z23.string(),
+  lineNumber: z23.number()
+});
+var DiffContentSchema = z23.object({
+  path: z23.string(),
+  expected: z23.string(),
+  actual: z23.string(),
+  diffLines: z23.array(DiffLineSchema)
+});
+var CommandExecutionSchema = z23.object({
+  command: z23.string(),
+  exitCode: z23.number(),
+  output: z23.string().optional(),
+  duration: z23.number()
+});
+var FileModificationSchema = z23.object({
+  path: z23.string(),
+  action: z23.enum(["created", "modified", "deleted"])
+});
+var ApiCallSchema = z23.object({
+  endpoint: z23.string(),
+  tokensUsed: z23.number(),
+  duration: z23.number()
+});
+var ExecutionTraceSchema = z23.object({
+  commands: z23.array(CommandExecutionSchema),
+  filesModified: z23.array(FileModificationSchema),
+  apiCalls: z23.array(ApiCallSchema),
+  totalDuration: z23.number()
+});
+var FailureAnalysisSchema = z23.object({
+  category: z23.enum(FailureCategory),
+  severity: z23.enum(FailureSeverity),
+  summary: z23.string(),
+  details: z23.string(),
+  rootCause: z23.string(),
+  suggestedFix: z23.string(),
+  relatedAssertions: z23.array(z23.string()),
+  codeSnippet: z23.string().optional(),
+  similarIssues: z23.array(z23.string()).optional(),
+  patternId: z23.string().optional(),
   // Extended fields for detailed debugging
   diff: DiffContentSchema.optional(),
   executionTrace: ExecutionTraceSchema.optional()
 });
 var EvalRunSchema = TenantEntitySchema.extend({
   /** Agent ID for this run */
-  agentId: z24.string().optional(),
+  agentId: z23.string().optional(),
   /** Skills group ID for this run */
-  skillsGroupId: z24.string().optional(),
+  skillsGroupId: z23.string().optional(),
   /** Scenario IDs to run */
-  scenarioIds: z24.array(z24.string()),
+  scenarioIds: z23.array(z23.string()),
   /** Current status */
   status: EvalStatusSchema,
   /** Progress percentage (0-100) */
-  progress: z24.number(),
+  progress: z23.number(),
   /** Results for each scenario/target combination */
-  results: z24.array(EvalRunResultSchema),
+  results: z23.array(EvalRunResultSchema),
   /** Aggregated metrics across all results */
   aggregateMetrics: EvalMetricsSchema,
   /** Failure analyses */
-  failureAnalyses: z24.array(FailureAnalysisSchema).optional(),
+  failureAnalyses: z23.array(FailureAnalysisSchema).optional(),
   /** Aggregated LLM trace summary */
   llmTraceSummary: LLMTraceSummarySchema.optional(),
   /** What triggered this run */
   trigger: TriggerSchema.optional(),
   /** When the run started (set when evaluation is triggered) */
-  startedAt: z24.string().optional(),
+  startedAt: z23.string().optional(),
   /** When the run completed */
-  completedAt: z24.string().optional(),
+  completedAt: z23.string().optional(),
   /** Live trace events captured during execution (for playback on results page) */
-  liveTraceEvents: z24.array(LiveTraceEventSchema).optional(),
+  liveTraceEvents: z23.array(LiveTraceEventSchema).optional(),
   /** Remote job ID for tracking execution in Dev Machines */
-  jobId: z24.string().optional(),
+  jobId: z23.string().optional(),
   /** Remote job status from the Dev Machine API (PENDING, RUNNING, COMPLETED, FAILED, CANCELLED) */
-  jobStatus: z24.string().optional(),
+  jobStatus: z23.string().optional(),
   /** Remote job error message if the job failed */
-  jobError: z24.string().optional(),
+  jobError: z23.string().optional(),
   /** Timestamp of the last job status check */
-  jobStatusCheckedAt: z24.string().optional()
+  jobStatusCheckedAt: z23.string().optional()
 });
 var CreateEvalRunInputSchema = EvalRunSchema.omit({
   id: true,
@@ -785,32 +703,119 @@ var CreateEvalRunInputSchema = EvalRunSchema.omit({
   startedAt: true,
   completedAt: true
 });
-var EvaluationProgressSchema = z24.object({
-  runId: z24.string(),
-  targetId: z24.string(),
-  totalScenarios: z24.number(),
-  completedScenarios: z24.number(),
-  scenarioProgress: z24.array(
-    z24.object({
-      scenarioId: z24.string(),
-      currentStep: z24.string(),
-      error: z24.string().optional()
+var EvaluationProgressSchema = z23.object({
+  runId: z23.string(),
+  targetId: z23.string(),
+  totalScenarios: z23.number(),
+  completedScenarios: z23.number(),
+  scenarioProgress: z23.array(
+    z23.object({
+      scenarioId: z23.string(),
+      currentStep: z23.string(),
+      error: z23.string().optional()
     })
   ),
-  createdAt: z24.number()
+  createdAt: z23.number()
 });
-var EvaluationLogSchema = z24.object({
-  runId: z24.string(),
-  scenarioId: z24.string(),
-  log: z24.object({
-    level: z24.enum(["info", "error", "debug"]),
-    message: z24.string().optional(),
-    args: z24.array(z24.any()).optional(),
-    error: z24.string().optional()
+var EvaluationLogSchema = z23.object({
+  runId: z23.string(),
+  scenarioId: z23.string(),
+  log: z23.object({
+    level: z23.enum(["info", "error", "debug"]),
+    message: z23.string().optional(),
+    args: z23.array(z23.any()).optional(),
+    error: z23.string().optional()
   })
 });
 var LLM_TIMEOUT = 12e4;
 
+// src/evaluation/eval-result.ts
+var AssertionResultStatus = /* @__PURE__ */ ((AssertionResultStatus2) => {
+  AssertionResultStatus2["PASSED"] = "passed";
+  AssertionResultStatus2["FAILED"] = "failed";
+  AssertionResultStatus2["SKIPPED"] = "skipped";
+  AssertionResultStatus2["ERROR"] = "error";
+  return AssertionResultStatus2;
+})(AssertionResultStatus || {});
+var AssertionResultSchema = z24.object({
+  id: z24.string(),
+  assertionId: z24.string(),
+  assertionType: z24.string(),
+  assertionName: z24.string(),
+  status: z24.enum(AssertionResultStatus),
+  message: z24.string().optional(),
+  expected: z24.string().optional(),
+  actual: z24.string().optional(),
+  duration: z24.number().optional(),
+  details: z24.record(z24.string(), z24.unknown()).optional(),
+  llmTraceSteps: z24.array(LLMTraceStepSchema).optional()
+});
+var EvalRunResultSchema = z24.object({
+  id: z24.string(),
+  targetId: z24.string(),
+  targetName: z24.string().optional(),
+  scenarioId: z24.string(),
+  scenarioName: z24.string(),
+  modelConfig: ModelConfigSchema.optional(),
+  assertionResults: z24.array(AssertionResultSchema),
+  metrics: EvalMetricsSchema.optional(),
+  passed: z24.number(),
+  failed: z24.number(),
+  passRate: z24.number(),
+  duration: z24.number(),
+  outputText: z24.string().optional(),
+  files: z24.array(ExpectedFileSchema).optional(),
+  /** File diffs showing changes made by the agent during execution */
+  fileDiffs: z24.array(DiffContentSchema).optional(),
+  startedAt: z24.string().optional(),
+  completedAt: z24.string().optional(),
+  llmTrace: LLMTraceSchema.optional()
+});
+var PromptResultSchema = z24.object({
+  text: z24.string(),
+  files: z24.array(z24.unknown()).optional(),
+  finishReason: z24.string().optional(),
+  reasoning: z24.string().optional(),
+  reasoningDetails: z24.unknown().optional(),
+  toolCalls: z24.array(z24.unknown()).optional(),
+  toolResults: z24.array(z24.unknown()).optional(),
+  warnings: z24.array(z24.unknown()).optional(),
+  sources: z24.array(z24.unknown()).optional(),
+  steps: z24.array(z24.unknown()),
+  generationTimeMs: z24.number(),
+  prompt: z24.string(),
+  systemPrompt: z24.string(),
+  usage: z24.object({
+    totalTokens: z24.number().optional(),
+    totalMicrocentsSpent: z24.number().optional()
+  })
+});
+var EvaluationResultSchema = z24.object({
+  id: z24.string(),
+  runId: z24.string(),
+  timestamp: z24.number(),
+  promptResult: PromptResultSchema,
+  testResults: z24.array(z24.unknown()),
+  tags: z24.array(z24.string()).optional(),
+  feedback: z24.string().optional(),
+  score: z24.number(),
+  suiteId: z24.string().optional()
+});
+var LeanEvaluationResultSchema = z24.object({
+  id: z24.string(),
+  runId: z24.string(),
+  timestamp: z24.number(),
+  tags: z24.array(z24.string()).optional(),
+  scenarioId: z24.string(),
+  scenarioVersion: z24.number().optional(),
+  targetId: z24.string(),
+  targetVersion: z24.number().optional(),
+  suiteId: z24.string().optional(),
+  score: z24.number(),
+  time: z24.number().optional(),
+  microcentsSpent: z24.number().optional()
+});
+
 // src/project/project.ts
 import { z as z25 } from "zod";
 var ProjectSchema = BaseEntitySchema.extend({