@langwatch/scenario 0.2.12 → 0.3.0

package/dist/index.mjs

@@ -17,11 +17,11 @@ import {
   getBatchRunId,
   getProjectConfig,
   scenarioProjectConfigSchema
-} from "./chunk-7HLDX5EL.mjs";
+} from "./chunk-3Z7E24UI.mjs";
 import {
   Logger,
   getEnv
-} from "./chunk-OL4RFXV4.mjs";
+} from "./chunk-RHTLQKEJ.mjs";
 import {
   __export
 } from "./chunk-7P6ASYW6.mjs";
@@ -35,7 +35,7 @@ __export(agents_exports, {
 
 // src/agents/judge-agent.ts
 import { generateText, tool } from "ai";
-import { z } from "zod";
+import { z } from "zod/v4";
 
 // src/agents/utils.ts
 var toolMessageRole = "tool";
@@ -142,14 +142,14 @@ ${criteriaList}
 function buildContinueTestTool() {
   return tool({
     description: "Continue the test with the next step",
-    parameters: z.object({})
+    inputSchema: z.object({})
   });
 }
 function buildFinishTestTool(criteria) {
   const criteriaNames = criteria.map(criterionToParamName);
   return tool({
     description: "Complete the test with a final verdict",
-    parameters: z.object({
+    inputSchema: z.object({
       criteria: z.object(
         Object.fromEntries(
           criteriaNames.map((name, idx) => [
@@ -207,7 +207,7 @@ var JudgeAgent = class extends JudgeAgentAdapter {
       model: mergedConfig.model,
       messages,
       temperature: mergedConfig.temperature ?? 0,
-      maxTokens: mergedConfig.maxTokens,
+      maxOutputTokens: mergedConfig.maxTokens,
       tools,
       toolChoice
     });
@@ -216,7 +216,7 @@ var JudgeAgent = class extends JudgeAgentAdapter {
       const toolCall = completion.toolCalls[0];
       switch (toolCall.toolName) {
         case "finish_test": {
-          args = toolCall.args;
+          args = toolCall.input;
           const verdict = args.verdict || "inconclusive";
           const reasoning = args.reasoning || "No reasoning provided";
           const criteria = args.criteria || {};
@@ -314,7 +314,7 @@ var UserSimulatorAgent = class extends UserSimulatorAgentAdapter {
       model: mergedConfig.model,
       messages: reversedMessages,
       temperature: mergedConfig.temperature ?? DEFAULT_TEMPERATURE,
-      maxTokens: mergedConfig.maxTokens
+      maxOutputTokens: mergedConfig.maxTokens
     });
     const messageContent = completion.text;
     if (!messageContent) {
@@ -441,9 +441,9 @@ var ScenarioExecutionState = class {
 };
 
 // src/utils/convert-core-messages-to-agui-messages.ts
-function convertCoreMessagesToAguiMessages(coreMessages) {
+function convertModelMessagesToAguiMessages(modelMessages) {
   const aguiMessages = [];
-  for (const msg of coreMessages) {
+  for (const msg of modelMessages) {
     const id = "id" in msg && typeof msg.id === "string" ? msg.id : generateMessageId();
     switch (true) {
       case msg.role === "system":
@@ -487,7 +487,7 @@ function convertCoreMessagesToAguiMessages(coreMessages) {
             type: "function",
             function: {
               name: c.toolName,
-              arguments: JSON.stringify(c.args)
+              arguments: JSON.stringify(c.input)
             }
           }))
         });
@@ -495,11 +495,12 @@ function convertCoreMessagesToAguiMessages(coreMessages) {
       }
       case msg.role === "tool":
         msg.content.map((p, i) => {
+          var _a;
           aguiMessages.push({
             id: `${id}-${i}`,
             role: "tool",
             toolCallId: p.toolCallId,
-            content: JSON.stringify(p.result)
+            content: JSON.stringify((_a = p.output) == null ? void 0 : _a.value)
           });
         });
         break;
@@ -509,7 +510,7 @@ function convertCoreMessagesToAguiMessages(coreMessages) {
   }
   return aguiMessages;
 }
-var convert_core_messages_to_agui_messages_default = convertCoreMessagesToAguiMessages;
+var convert_core_messages_to_agui_messages_default = convertModelMessagesToAguiMessages;
 
 // src/execution/scenario-execution.ts
 var ScenarioExecution = class {
@@ -575,7 +576,7 @@ var ScenarioExecution = class {
   /**
    * Gets the complete conversation history as an array of messages.
    *
-   * @returns Array of CoreMessage objects representing the full conversation
+   * @returns Array of ModelMessage objects representing the full conversation
    */
   get messages() {
     return this.state.messages;
@@ -808,7 +809,7 @@ var ScenarioExecution = class {
    * - "assistant" messages are routed to AGENT role agents
    * - Other message types are added directly to the conversation
    *
-   * @param message - The CoreMessage to add to the conversation
+   * @param message - The ModelMessage to add to the conversation
    *
    * @example
    * ```typescript
@@ -837,7 +838,7 @@ var ScenarioExecution = class {
    *
    * This method is part of the ScenarioExecutionLike interface used by script steps.
    *
-   * @param content - Optional content for the user's message. Can be a string or CoreMessage.
+   * @param content - Optional content for the user's message. Can be a string or ModelMessage.
    *                 If not provided, the user simulator agent will generate the content.
    *
    * @example
@@ -848,7 +849,7 @@ var ScenarioExecution = class {
    * // Let user simulator generate content
    * await execution.user();
    *
-   * // Use a CoreMessage object
+   * // Use a ModelMessage object
    * await execution.user({
    *   role: "user",
    *   content: "Tell me a joke"
@@ -867,7 +868,7 @@ var ScenarioExecution = class {
    *
    * This method is part of the ScenarioExecutionLike interface used by script steps.
    *
-   * @param content - Optional content for the agent's response. Can be a string or CoreMessage.
+   * @param content - Optional content for the agent's response. Can be a string or ModelMessage.
    *                 If not provided, the agent under test will generate the response.
    *
    * @example
@@ -878,7 +879,7 @@ var ScenarioExecution = class {
    * // Use provided content
    * await execution.agent("The weather is sunny today!");
    *
-   * // Use a CoreMessage object
+   * // Use a ModelMessage object
    * await execution.agent({
    *   role: "assistant",
    *   content: "I'm here to help you with weather information."
@@ -1600,14 +1601,13 @@ function formatPart(part) {
     case "file":
       return `(file): ${part.filename} ${typeof part.data === "string" ? `url:${part.data}` : "base64:omitted"}`;
     case "tool-call":
-      return `(tool call): ${part.toolName} id:${part.toolCallId} args:(${JSON.stringify(part.args)})`;
+      return `(tool call): ${part.toolName} id:${part.toolCallId} args:(${JSON.stringify(part.input)})`;
     case "tool-result":
-      return `(tool result): ${part.toolName} id:${part.toolCallId} result:(${JSON.stringify(part.result)})`;
+      return `(tool result): ${part.toolName} id:${part.toolCallId} result:(${JSON.stringify(part.output)})`;
     case "reasoning":
       return `(reasoning): ${part.text}`;
-    case "redacted-reasoning":
-      return `(redacted reasoning): ${part.data}`;
     default:
+      part;
       return `Unknown content: ${JSON.stringify(part)}`;
   }
 }

package/dist/integrations/vitest/reporter.js

@@ -38,7 +38,7 @@ var import_path = __toESM(require("path"));
 var import_chalk = __toESM(require("chalk"));
 
 // src/config/env.ts
-var import_zod = require("zod");
+var import_v4 = require("zod/v4");
 
 // src/config/log-levels.ts
 var LogLevel = /* @__PURE__ */ ((LogLevel2) => {
@@ -51,37 +51,37 @@ var LogLevel = /* @__PURE__ */ ((LogLevel2) => {
 var LOG_LEVELS = Object.values(LogLevel);
 
 // src/config/env.ts
-var envSchema = import_zod.z.object({
+var envSchema = import_v4.z.object({
   /**
    * LangWatch API key for event reporting.
    * If not provided, events will not be sent to LangWatch.
    */
-  LANGWATCH_API_KEY: import_zod.z.string().optional(),
+  LANGWATCH_API_KEY: import_v4.z.string().optional(),
   /**
    * LangWatch endpoint URL for event reporting.
    * Defaults to the production LangWatch endpoint.
    */
-  LANGWATCH_ENDPOINT: import_zod.z.string().url().optional().default("https://app.langwatch.ai"),
+  LANGWATCH_ENDPOINT: import_v4.z.string().url().optional().default("https://app.langwatch.ai"),
   /**
    * Disables simulation report info messages when set to any truthy value.
    * Useful for CI/CD environments or when you want cleaner output.
    */
-  SCENARIO_DISABLE_SIMULATION_REPORT_INFO: import_zod.z.string().optional().transform((val) => Boolean(val)),
+  SCENARIO_DISABLE_SIMULATION_REPORT_INFO: import_v4.z.string().optional().transform((val) => Boolean(val)),
   /**
    * Node environment - affects logging and behavior.
    * Defaults to 'development' if not specified.
    */
-  NODE_ENV: import_zod.z.enum(["development", "production", "test"]).default("development"),
+  NODE_ENV: import_v4.z.enum(["development", "production", "test"]).default("development"),
   /**
    * Case-insensitive log level for the scenario package.
    * Defaults to 'info' if not specified.
    */
-  LOG_LEVEL: import_zod.z.string().toUpperCase().pipe(import_zod.z.nativeEnum(LogLevel)).optional().default("INFO" /* INFO */),
+  LOG_LEVEL: import_v4.z.string().toUpperCase().pipe(import_v4.z.nativeEnum(LogLevel)).optional().default("INFO" /* INFO */),
   /**
    * Scenario batch run ID.
    * If not provided, a random ID will be generated.
    */
-  SCENARIO_BATCH_RUN_ID: import_zod.z.string().optional()
+  SCENARIO_BATCH_RUN_ID: import_v4.z.string().optional()
 });
 function getEnv() {
   return envSchema.parse(process.env);
@@ -321,8 +321,12 @@ ${indent(parsedJson)}
     console.log();
     console.log(import_chalk.default.bold.cyan("=== Scenario Test Report ==="));
     console.log(`Total Scenarios: ${total}`);
-    console.log(import_chalk.default.green(`Passed: ${passed}`));
-    console.log(import_chalk.default.red(`Failed: ${failed}`));
+    console.log(
+      passed > 0 ? import_chalk.default.green(`Passed: ${passed}`) : `Passed: ${passed}`
+    );
+    console.log(
+      failed > 0 ? import_chalk.default.red(`Failed: ${failed}`) : `Failed: ${failed}`
+    );
     console.log(`Success Rate: ${import_chalk.default.bold(`${successRate}%`)}`);
     this.results.forEach((r, i) => {
       const statusColor = r.status === "SUCCESS" ? import_chalk.default.green : import_chalk.default.red;

package/dist/integrations/vitest/reporter.mjs

@@ -1,6 +1,6 @@
 import {
   Logger
-} from "../../chunk-OL4RFXV4.mjs";
+} from "../../chunk-RHTLQKEJ.mjs";
 import "../../chunk-7P6ASYW6.mjs";
 
 // src/integrations/vitest/reporter.ts
@@ -161,8 +161,12 @@ ${indent(parsedJson)}
     console.log();
     console.log(chalk.bold.cyan("=== Scenario Test Report ==="));
     console.log(`Total Scenarios: ${total}`);
-    console.log(chalk.green(`Passed: ${passed}`));
-    console.log(chalk.red(`Failed: ${failed}`));
+    console.log(
+      passed > 0 ? chalk.green(`Passed: ${passed}`) : `Passed: ${passed}`
+    );
+    console.log(
+      failed > 0 ? chalk.red(`Failed: ${failed}`) : `Failed: ${failed}`
+    );
     console.log(`Success Rate: ${chalk.bold(`${successRate}%`)}`);
     this.results.forEach((r, i) => {
       const statusColor = r.status === "SUCCESS" ? chalk.green : chalk.red;

package/dist/integrations/vitest/setup-global.js

@@ -25,6 +25,6 @@ __export(setup_global_exports, {
 module.exports = __toCommonJS(setup_global_exports);
 var import_xksuid = require("xksuid");
 function setup() {
-  const scenarioBatchRunId = `scenariobatchrun_${(0, import_xksuid.generate)()}`;
+  const scenarioBatchRunId = `scenariobatch_${(0, import_xksuid.generate)()}`;
   process.env.SCENARIO_BATCH_RUN_ID = scenarioBatchRunId;
 }

package/dist/integrations/vitest/setup-global.mjs

@@ -3,7 +3,7 @@ import "../../chunk-7P6ASYW6.mjs";
 // src/integrations/vitest/setup-global.ts
 import { generate } from "xksuid";
 function setup() {
-  const scenarioBatchRunId = `scenariobatchrun_${generate()}`;
+  const scenarioBatchRunId = `scenariobatch_${generate()}`;
   process.env.SCENARIO_BATCH_RUN_ID = scenarioBatchRunId;
 }
 export {