@wix/evalforge-types 0.34.0 → 0.36.0

package/build/index.js

@@ -915,6 +915,9 @@ var require_types_impl = __commonJS({
 var index_exports = {};
 __export(index_exports, {
   AVAILABLE_MODEL_IDS: () => AVAILABLE_MODEL_IDS,
+  AVAILABLE_RUN_COMMANDS: () => AVAILABLE_RUN_COMMANDS,
+  AgentRunCommand: () => AgentRunCommand,
+  AgentRunCommandSchema: () => AgentRunCommandSchema,
   AgentSchema: () => AgentSchema,
   AllowedCommands: () => AllowedCommands,
   ApiCallSchema: () => ApiCallSchema,
@@ -992,6 +995,7 @@ __export(index_exports, {
   PlaywrightNLTestSchema: () => PlaywrightNLTestSchema,
   ProjectSchema: () => ProjectSchema,
   PromptResultSchema: () => PromptResultSchema,
+  RUN_COMMAND_LABELS: () => RUN_COMMAND_LABELS,
   SEMVER_REGEX: () => SEMVER_REGEX,
   SKILL_FOLDER_NAME_REGEX: () => SKILL_FOLDER_NAME_REGEX,
   SYSTEM_ASSERTIONS: () => SYSTEM_ASSERTIONS,
@@ -1021,6 +1025,8 @@ __export(index_exports, {
   TestSuiteSchema: () => TestSuiteSchema,
   TestType: () => TestType,
   TestTypeSchema: () => TestTypeSchema,
+  TimeAssertionSchema: () => TimeAssertionSchema,
+  TimeConfigSchema: () => TimeConfigSchema,
   TokenUsageSchema: () => TokenUsageSchema,
   ToolTestSchema: () => ToolTestSchema,
   TriggerMetadataSchema: () => TriggerMetadataSchema,
@@ -1128,9 +1134,18 @@ var TargetSchema = TenantEntitySchema.extend({
 
 // src/target/agent.ts
 var import_zod5 = require("zod");
+var AgentRunCommand = /* @__PURE__ */ ((AgentRunCommand2) => {
+  AgentRunCommand2["CLAUDE"] = "claude";
+  return AgentRunCommand2;
+})(AgentRunCommand || {});
+var AVAILABLE_RUN_COMMANDS = Object.values(AgentRunCommand);
+var RUN_COMMAND_LABELS = {
+  ["claude" /* CLAUDE */]: "Claude Code"
+};
+var AgentRunCommandSchema = import_zod5.z.nativeEnum(AgentRunCommand);
 var AgentSchema = TargetSchema.extend({
   /** Command to run the agent */
-  runCommand: import_zod5.z.string(),
+  runCommand: AgentRunCommandSchema,
   /** Optional model configuration override */
   modelConfig: ModelConfigSchema.optional()
 });
@@ -1469,9 +1484,15 @@ var LlmJudgeAssertionSchema = import_zod20.z.object({
   maxTokens: import_zod20.z.number().int().optional(),
   temperature: import_zod20.z.number().min(0).max(1).optional()
 });
+var TimeAssertionSchema = import_zod20.z.object({
+  type: import_zod20.z.literal("time_limit"),
+  /** Maximum allowed duration in milliseconds */
+  maxDurationMs: import_zod20.z.number().int().positive()
+});
 var AssertionSchema = import_zod20.z.union([
   SkillWasCalledAssertionSchema,
   BuildPassedAssertionSchema,
+  TimeAssertionSchema,
   LlmJudgeAssertionSchema
 ]);
 
@@ -1517,6 +1538,7 @@ var import_zod22 = require("zod");
 var AssertionTypeSchema = import_zod22.z.enum([
   "skill_was_called",
   "build_passed",
+  "time_limit",
   "llm_judge"
 ]);
 var AssertionParameterTypeSchema = import_zod22.z.enum([
@@ -1557,6 +1579,10 @@ var BuildPassedConfigSchema = import_zod22.z.strictObject({
   /** Expected exit code (default: 0) */
   expectedExitCode: import_zod22.z.number().int().optional()
 });
+var TimeConfigSchema = import_zod22.z.strictObject({
+  /** Maximum allowed duration in milliseconds */
+  maxDurationMs: import_zod22.z.number().int().positive()
+});
 var LlmJudgeConfigSchema = import_zod22.z.object({
   /**
    * Prompt template with placeholders:
@@ -1586,7 +1612,9 @@ var AssertionConfigSchema = import_zod22.z.union([
   LlmJudgeConfigSchema,
   // requires prompt - check first
   SkillWasCalledConfigSchema,
-  // requires skillName
+  // requires skillNames
+  TimeConfigSchema,
+  // requires maxDurationMs, uses strictObject
   BuildPassedConfigSchema,
   // all optional, uses strictObject to reject unknown keys
   import_zod22.z.object({})
@@ -1611,6 +1639,8 @@ function validateAssertionConfig(type, config) {
       return SkillWasCalledConfigSchema.safeParse(config).success;
     case "build_passed":
       return BuildPassedConfigSchema.safeParse(config).success;
+    case "time_limit":
+      return TimeConfigSchema.safeParse(config).success;
     case "llm_judge":
       return LlmJudgeConfigSchema.safeParse(config).success;
     default:
@@ -2114,6 +2144,7 @@ var UpdateTemplateInputSchema = CreateTemplateInputSchema.partial();
 var SYSTEM_ASSERTION_IDS = {
   SKILL_WAS_CALLED: "system:skill_was_called",
   BUILD_PASSED: "system:build_passed",
+  TIME_LIMIT: "system:time_limit",
   LLM_JUDGE: "system:llm_judge"
 };
 function isSystemAssertionId(id) {
@@ -2170,6 +2201,21 @@ var SYSTEM_ASSERTIONS = {
       }
     ]
   },
+  [SYSTEM_ASSERTION_IDS.TIME_LIMIT]: {
+    id: SYSTEM_ASSERTION_IDS.TIME_LIMIT,
+    name: "Time Limit",
+    description: "Check that the scenario completed within a maximum duration",
+    type: "time_limit",
+    parameters: [
+      {
+        name: "maxDurationMs",
+        label: "Max Duration (ms)",
+        type: "number",
+        required: true,
+        defaultValue: 3e5
+      }
+    ]
+  },
   [SYSTEM_ASSERTION_IDS.LLM_JUDGE]: {
     id: SYSTEM_ASSERTION_IDS.LLM_JUDGE,
     name: "LLM Judge",
@@ -2221,6 +2267,9 @@ function getSystemAssertion(id) {
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {
   AVAILABLE_MODEL_IDS,
+  AVAILABLE_RUN_COMMANDS,
+  AgentRunCommand,
+  AgentRunCommandSchema,
   AgentSchema,
   AllowedCommands,
   ApiCallSchema,
@@ -2298,6 +2347,7 @@ function getSystemAssertion(id) {
   PlaywrightNLTestSchema,
   ProjectSchema,
   PromptResultSchema,
+  RUN_COMMAND_LABELS,
   SEMVER_REGEX,
   SKILL_FOLDER_NAME_REGEX,
   SYSTEM_ASSERTIONS,
@@ -2327,6 +2377,8 @@ function getSystemAssertion(id) {
   TestSuiteSchema,
   TestType,
   TestTypeSchema,
+  TimeAssertionSchema,
+  TimeConfigSchema,
   TokenUsageSchema,
   ToolTestSchema,
   TriggerMetadataSchema,