@empiricalrun/test-gen 0.47.3 → 0.47.4

package/dist/agent/codegen/update-flow.js

@@ -5,16 +5,14 @@ var __importDefault = (this && this.__importDefault) || function (mod) {
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.appendCreateTestBlock = exports.getAppendCreateTestBlockCompletion = exports.updateTest = exports.getUpdateTestCodeCompletion = void 0;
 const llm_1 = require("@empiricalrun/llm");
-const crypto_1 = __importDefault(require("crypto"));
 const fs_extra_1 = __importDefault(require("fs-extra"));
 const logger_1 = require("../../bin/logger");
 const context_1 = require("../../bin/utils/context");
 const fs_1 = require("../../bin/utils/fs");
 const web_1 = require("../../bin/utils/platform/web");
 const constants_1 = require("../../constants");
-const promptTemplate_0 = "{{#section \"system\"}}\n\nYou are a software test engineer who is given a task to update a test case. You will be provided with steps of a test\ncase scenario and you are given a snippet with interface await createTest(task, playwright_page_instance).\n\nYou need to analyze the request and place createTest snippet at the correct position and pass on the steps to this\n`createTest` method with the correct page instance.\n\nYou will be provided with current tests, fixtures and page object models for you to use and update code as per the\ntask provided to you.\nYou need to respond with file path and updated code block inside the file.\n\nHere is the list of current tests and fixtures:\n\n{{testFiles}}\n\n\nHere is the list of current page object models:\n\n{{pageFiles}}\n{{/section}}\n\n\n{{#section \"user\"}}\nFollowing is the test scenario for which you need to update the test:\ntest name:\n{{scenarioName}}\n\n\nTask:\n{{scenarioSteps}}\n\n\nFollow these steps to complete the task:\n\n1. Determine the task's intent:\n- The default intent is **\"add steps\"**.\n- If the task explicitly mentions (using words like \"replace\", \"remove\", \"delete\") that existing test steps should\nbe replaced or deleted, the intent is **\"update steps\"**. Identify the specific steps to be replaced or removed.\n- Note:\n- Do not assume that providing new or different steps implies an intent to update or replace existing steps. Unless\nthe task explicitly instructs to replace or remove existing code, interpret the intent as adding new steps to the\nexisting test.\n\n2. Identify the test block that requires updating.\n\n3. Place the `createTest` snippet:\n- Insert the `createTest` snippet at the location determined by the task, and pass the Playwright page instance.\n- If the intent is \"add\", do not alter the existing test code; simply append the `createTest` snippet based on any\nprovided location hints.\n- If the task includes location hints that don't match steps within the test, check dependent methods called from\nthe test for the update.\n- If no location hint is provided, place the `createTest` snippet at the end of the test block.\n- Even if the task includes steps that overlap with or are similar to existing steps, do not modify the existing\ncode. Unless the task explicitly instructs to replace or remove existing code, interpret the intent as adding new\nsteps to the existing test.\n\n4. Strip location hints from the task:\n- Remove any location hints (e.g., \"replace the current assertion and\") before passing the task to the `createTest`\nmethod.\n\nSubmission Guidelines:\n\n- Focus only on the provided test case and any related page object model methods used in the test.\n- Do not modify or add code within the `createTest` snippet.\n- Do not update or modify any other code apart from adding `createTest` code snippet.\n- Since the response will be used for search-and-replace operations, always provide the immediate parent AST node\nfor any code updates.\n- Include the full test block if any part of it is updated, preserving all unchanged code.\n- Do not use markdown syntax or backticks.\n- Respond using the following XML format:\n<reason_for_intent></reason_for_intent>\n<intent></intent>\n<location_of_update></location_of_update>\n<file_path></file_path>\n<old_code_block></old_code_block>\n<new_code_block></new_code_block>\n<change></change>\n\n- Each `<old_code_block>` and `<new_code_block>` should contain only one test block or page object model method\n    definition. Provide separate blocks for multiple updates.\n    - The `<change></change>` tag should also mention the file path being updated.\n    - There should be only one `createTest` block in the `new_code_block`. The `createTest` method should be passed\n    with entire task. Do not split the task while forwarding it to `createTest`.\n    - `<new_code_block>` code snippet should be syntactically correct.\n      - The code_block should not contain any import statements.\n      {{/section}}";
+const promptTemplate_0 = "{{#section \"system\"}}\n\nYou are a software test engineer who is given a task to update a test case. You will be provided with steps of a test\ncase scenario and you are given a snippet with interface await createTest(task, playwright_page_instance).\n\nYou need to analyze the request and place createTest snippet at the correct position and pass on the steps to this\n`createTest` method with the correct page instance.\n\nYou will be provided with current tests, fixtures and page object models for you to use and update code as per the\ntask provided to you.\nYou need to respond with file path and updated code block inside the file.\n\nHere is the list of current tests and fixtures:\n\n{{testFiles}}\n\n\nHere is the list of current page object models:\n\n{{pageFiles}}\n{{/section}}\n\n\n{{#section \"user\"}}\nFollowing is the test scenario for which you need to update the test:\n\ntest name:\n{{scenarioName}}\n\nTask:\n{{scenarioSteps}}\n\nFollow these steps to complete the task:\n\n1. Determine the task's intent:\n- The default intent is **\"add steps\"**.\n- If the task explicitly mentions (using words like \"replace\", \"remove\", \"delete\") that existing test steps should\nbe replaced or deleted, the intent is **\"update steps\"**. Identify the specific steps to be replaced or removed.\n- Do not assume that providing new or different steps implies an intent to update or replace existing steps. Unless\nthe task explicitly instructs to replace or remove existing code, interpret the intent as adding new steps to the\nexisting test.\n\n2. Identify the test block that requires updating.\n\n3. Place the `createTest` snippet:\n- Insert the `createTest` snippet at the location determined by the task, and pass the Playwright page instance.\n- If the intent is \"add\", do not alter the existing test code; simply append the `createTest` snippet based on any\nprovided location hints.\n- If the task includes location hints that don't match steps within the test, check dependent methods called from\nthe test for the update.\n- If no location hint is provided, place the `createTest` snippet at the end of the test block.\n- Even if the task includes steps that overlap with or are similar to existing steps, do not modify the existing\ncode. Unless the task explicitly instructs to replace or remove existing code, interpret the intent as adding new\nsteps to the existing test.\n\n4. Strip location hints from the task:\n- Remove any location hints (e.g., \"replace the current assertion and\") before passing the task to the `createTest`\nmethod.\n- The task passed into `createTest` should not contain any location hints. It should only contain actions that\nare required to be performed.\n- For example, if the original task: \"replace the current assertion and add a new assertion\" should be transformed to\n\"add a new assertion\" before passing it to the `createTest` method. This is the \"task_without_location_hints\"\n\nSubmission Guidelines:\n\n- Focus only on the provided test case and any related page object model methods used in the test.\n- Do not modify or add code within the `createTest` snippet.\n- Do not update or modify any other code apart from adding `createTest` code snippet.\n- Since the response will be used for search-and-replace operations, always provide the immediate parent AST node\nfor any code updates.\n- Include the full test block if any part of it is updated, preserving all unchanged code.\n- Do not use markdown syntax or backticks.\n- Respond using the following XML format:\n<reason_for_intent></reason_for_intent>\n<intent></intent>\n<location_of_update></location_of_update>\n<task_without_location_hints></task_without_location_hints>\n<file_path></file_path>\n<old_code_block></old_code_block>\n<new_code_block></new_code_block>\n<change></change>\n\n- Each `<old_code_block>` and `<new_code_block>` should contain only one test block or page object model method\n  definition. Provide separate blocks for multiple updates.\n- The `<change></change>` tag should also mention the file path being updated.\n- There should be only one `createTest` block in the `new_code_block`. The `createTest` method should be passed\nwith entire task. Do not split the task while forwarding it to `createTest`.\n- `<new_code_block>` code snippet should be syntactically correct.\n  - The code_block should not contain any import statements.\n{{/section}}";
 const promptTemplate_1 = "{{#section \"system\"}}\nYou are a software test engineer who is given an objective to update test basis the task provided.\nYou will be provided with a test name, test and test file path.\n\nYou will be provided with current tests, fixtures and page object models for you to use and update code as per the task\nprovided to you. You need to respond with file path and updated code block inside the file.\n\nBefore responding you need to ensure that the code change is minimal and the change is reusable across tests. You need\nto ensure the code follows DRY principle.\n\nHere is the list of current tests and fixtures:\n\n{{testFiles}}\n\n\nHere is the list of current page object models:\n\n{{pageFiles}}\n\n{{/section}}\n\n{{#section \"user\"}}\nFollowing are the test details and the task to complete your objective:\nTest name:\n{{scenarioName}}\n\nTest case:\n{{currentScenarioCodeBlock}}\n\nTask:\n{{scenarioSteps}}\n\nIn order to execute the task:\n- Think step by step and first identify current test block which needs update and the methods which the current test\nblock depend on.\n- The task will demand changes in the current test case or the methods it depend on. Based on the task, identify list of\nfile paths which need change, the reason for change and the code change they need.\n- Next make changes to code blocks in each file paths.\n\n------\n\nFollow these instructions before responding with output:\n- Read the code line by line and ensure that achieve the task provided to you\n- Read the dependencies of the code block by scanning through file paths and file provided to you. refer the same file\npath while responding with output.\n- Focus only on the test case provided and associated JS methods called from the test case.\n- Since the response will be used to search and replace blocks, always respond with output which includes the full\nlexical scope surrounding the modified code.\n- If there are any updates inside test code block, ensure responding with full test block with unchanged code as well\n- Each code block should contain edits to only one code block in file path\n- DO NOT respond with any backticks or markdown syntax\n- Respond only with file path where the code block to be updated is present, old code block, new code block and a one\nliner reason for the change\n- Respond with <file_path></file_path>, <old_code_block></old_code_block>, <new_code_block></new_code_block> and\n<change></change> as xml tags\n- The reason for change should adhere to coding principles provided and review if the updated code is present in the\nfile path mentioned\n- The code change should belong to the right file path\n- The response must start with <file_path>\n  {{/section}}";
-const session_1 = require("../../session");
 const test_update_feedback_1 = require("./test-update-feedback");
 const utils_1 = require("./utils");
 async function getUpdateTestCodeCompletion({ testCase, testFileContent, testFiles, pageFiles, testFilePath, options, trace, }) {
@@ -65,31 +63,18 @@ async function getUpdateTestCodeCompletion({ testCase, testFileContent, testFile
     };
 }
 exports.getUpdateTestCodeCompletion = getUpdateTestCodeCompletion;
-async function updateTest(testCase, file, options, logging = true, validate = true, trace) {
-    const logger = new logger_1.CustomLogger({ useReporter: logging });
+async function updateTest({ testCase, file, trace, }) {
+    const logger = new logger_1.CustomLogger({ useReporter: true });
     const context = await (0, context_1.contextForGeneration)(file);
     const { codePrompt, pomPrompt, nonSpecFilePrompt, testFileContent } = context;
     const generatedTestCases = [];
     logger.logEmptyLine();
-    const session = (0, session_1.getSessionDetails)();
-    trace =
-        trace ||
-            llm_1.langfuseInstance?.trace({
-                name: "update-test",
-                id: crypto_1.default.randomUUID(),
-                release: session.version,
-                tags: [
-                    options?.metadata.projectName || "",
-                    options?.metadata.environment || "",
-                ].filter((s) => !!s),
-            });
     const request = {
         testCase,
         testFileContent,
         testFiles: codePrompt,
         pageFiles: pomPrompt,
         testFilePath: file,
-        options,
     };
     const updateTestSpan = trace?.span({
         name: "update-test",
@@ -105,7 +90,6 @@ async function updateTest(testCase, file, options, logging = true, validate = tr
         testCase,
         fileChanges,
         logger,
-        testGenOptions: options,
         pomPrompt: pomPrompt,
         nonSpecFilePrompt: nonSpecFilePrompt,
     });
@@ -125,18 +109,16 @@ async function updateTest(testCase, file, options, logging = true, validate = tr
             testCase,
             fileChanges: filteredFileChanges,
             logger,
-            testGenOptions: options,
             pomPrompt: pomPrompt,
             nonSpecFilePrompt: nonSpecFilePrompt,
         });
     }
     await (0, utils_1.validateTypesAndFormatCode)({
-        validateTypes: validate,
+        validateTypes: true,
         trace: updateTestSpan,
         testCase,
         filePaths: fileChanges.map((f) => f.filePath),
         logger,
-        testGenOptions: options,
         pomPrompt: pomPrompt,
         nonSpecFilePrompt: nonSpecFilePrompt,
     });
@@ -153,19 +135,7 @@ async function updateTest(testCase, file, options, logging = true, validate = tr
     return generatedTestCases;
 }
 exports.updateTest = updateTest;
-async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCase, testFilePath, options, trace, }) {
-    const session = (0, session_1.getSessionDetails)();
-    trace =
-        trace ||
-            llm_1.langfuseInstance?.trace({
-                name: "append-create-test-block",
-                id: crypto_1.default.randomUUID(),
-                release: session.version,
-                tags: [
-                    options?.metadata.projectName || "",
-                    options?.metadata.environment || "",
-                ].filter((s) => !!s),
-            });
+async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCase, testFilePath, trace, }) {
     const appendCreateTestSpan = trace?.span({
         name: "append-create-test-block",
         input: {
@@ -175,9 +145,6 @@ async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCa
             testFilePath,
         },
     });
-    const promptSpan = trace?.span({
-        name: "append-create-test-block-prompt",
-    });
     const instruction = (0, llm_1.compilePrompt)(promptTemplate_0, {
         testFiles: testFiles,
         pageFiles: pageFiles,
@@ -185,7 +152,6 @@ async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCa
         scenarioSteps: testCase.steps.join("\n"),
         scenarioFile: testFilePath,
     });
-    promptSpan?.end({ output: { instruction } });
     const [userInstruction] = instruction.filter((s) => s.role === "user");
     const [systemInstruction] = instruction.filter((s) => s.role === "system");
     userInstruction.content = `${systemInstruction?.content}
@@ -193,7 +159,7 @@ async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCa
     ${userInstruction?.content}
   `;
     const llm = new llm_1.LLM({
-        trace,
+        trace: appendCreateTestSpan,
         provider: "anthropic",
         defaultModel: "claude-3-5-sonnet-latest",
         providerApiKey: constants_1.MODEL_API_KEYS["anthropic"],
@@ -201,7 +167,6 @@ async function getAppendCreateTestBlockCompletion({ testFiles, pageFiles, testCa
     const chatCompletionParams = {
         messages: [userInstruction],
         modelParameters: {
-            ...options?.modelParameters,
             ...constants_1.DEFAULT_MODEL_PARAMETERS,
         },
     };
@@ -245,7 +210,7 @@ Respond with the same xml format which was originally expected.
     return response;
 }
 exports.getAppendCreateTestBlockCompletion = getAppendCreateTestBlockCompletion;
-async function appendCreateTestBlock({ testCase, file, options, trace, validateTypes = true, }) {
+async function appendCreateTestBlock({ testCase, file, trace, validateTypes = true, }) {
     const logger = new logger_1.CustomLogger({ useReporter: false });
     logger.log("Appending create test block");
     const context = await (0, context_1.contextForGeneration)(file);
@@ -258,22 +223,19 @@ async function appendCreateTestBlock({ testCase, file, options, trace, validateT
     ]);
     const { pomPrompt, nonSpecFilePrompt } = context;
     const generatedTestCases = [];
-    const appendCreateTestParams = {
+    const appendCreateTestResp = await getAppendCreateTestBlockCompletion({
         testCase,
         testFilePath: file,
-        options,
         trace,
         pageFiles: pomPrompt,
         testFiles: testCodePrompt,
-    };
-    const appendCreateTestResp = await getAppendCreateTestBlockCompletion(appendCreateTestParams);
+    });
     const fileChanges = (0, utils_1.extractAppendTestUpdates)(appendCreateTestResp);
     await (0, utils_1.applyFileChanges)({
         trace,
         testCase,
         fileChanges,
         logger,
-        testGenOptions: options,
         pomPrompt: pomPrompt,
         nonSpecFilePrompt: nonSpecFilePrompt,
     });
@@ -282,7 +244,6 @@ async function appendCreateTestBlock({ testCase, file, options, trace, validateT
         testCase,
         filePaths: fileChanges.map((f) => f.filePath),
         logger,
-        testGenOptions: options,
         pomPrompt: pomPrompt,
         nonSpecFilePrompt: nonSpecFilePrompt,
         validateTypes,
@@ -294,7 +255,6 @@ async function appendCreateTestBlock({ testCase, file, options, trace, validateT
         ...testCase,
         updatedFiles: fileChanges.map((f) => f.filePath),
     });
-    await (0, llm_1.flushAllTraces)();
     return generatedTestCases;
 }
 exports.appendCreateTestBlock = appendCreateTestBlock;

package/dist/agent/codegen/utils.d.ts

@@ -1,5 +1,5 @@
 import { TraceClient } from "@empiricalrun/llm";
-import type { TestCase, TestGenConfigOptions } from "@empiricalrun/shared-types";
+import type { TestCase } from "@empiricalrun/shared-types";
 import { CustomLogger } from "../../bin/logger";
 import { TestGenUpdatesReporter } from "../../reporter";
 import { CodeUpdate, CreateTestCodeUpdate, FileUpdateResponse } from "./types";
@@ -57,13 +57,12 @@ export declare function extractTestStepsSuggestions(input: string): {
     reason: string;
     methodName: string;
 }[];
-export declare function validateTypesAndFormatCode({ validateTypes, trace, testCase, filePaths, logger, testGenOptions, pomPrompt, nonSpecFilePrompt, }: {
+export declare function validateTypesAndFormatCode({ validateTypes, trace, testCase, filePaths, logger, pomPrompt, nonSpecFilePrompt, }: {
     validateTypes?: boolean;
     trace?: TraceClient;
     testCase: TestCase;
     filePaths: (string | undefined)[];
     logger?: CustomLogger;
-    testGenOptions?: TestGenConfigOptions;
     pomPrompt?: string;
     nonSpecFilePrompt?: string;
 }): Promise<void>;
@@ -103,7 +102,6 @@ export declare function applyFileChanges({ trace, testCase, fileChanges, logger,
         reason: string | undefined;
     }[];
     logger?: CustomLogger;
-    testGenOptions?: TestGenConfigOptions;
     pomPrompt?: string;
     nonSpecFilePrompt?: string;
 }): Promise<{

package/dist/agent/codegen/utils.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"utils.d.ts","sourceRoot":"","sources":["../../../src/agent/codegen/utils.ts"],"names":[],"mappings":"AAAA,OAAO,EAAkB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AAChE,OAAO,KAAK,EACV,QAAQ,EACR,oBAAoB,EACrB,MAAM,4BAA4B,CAAC;AAIpC,OAAO,EAAE,YAAY,EAAE,MAAM,kBAAkB,CAAC;AAMhD,OAAO,EAAE,sBAAsB,EAAE,MAAM,gBAAgB,CAAC;AAExD,OAAO,EAAE,UAAU,EAAE,oBAAoB,EAAE,kBAAkB,EAAE,MAAM,SAAS,CAAC;AAE/E;;;;;;;;;;;GAWG;AACH,wBAAgB,kBAAkB,CAAC,KAAK,EAAE,MAAM,GAAG,UAAU,EAAE,CAiB9D;AAED;;;;;;;;;;GAUG;AACH,wBAAgB,wBAAwB,CACtC,KAAK,EAAE,MAAM,GACZ,oBAAoB,EAAE,CAgBxB;AAED;;;;;;;;;;;;;GAaG;AACH,wBAAgB,wBAAwB,CAAC,KAAK,EAAE,MAAM,GAAG;IACvD,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,cAAc,EAAE,MAAM,GAAG,SAAS,CAAC;IACnC,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;IAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;CAC5B,EAAE,CA8BF;AAED,wBAAgB,2BAA2B,CAAC,KAAK,EAAE,MAAM,GAAG;IAC1D,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,CAAC;IACjB,YAAY,EAAE,MAAM,CAAC;IACrB,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;CACpB,EAAE,CAqCF;AAED,wBAAsB,0BAA0B,CAAC,EAC/C,aAAoB,EACpB,KAAK,EACL,QAAQ,EACR,SAAS,EACT,MAAM,EACN,cAAc,EACd,SAAS,EACT,iBAAiB,GAClB,EAAE;IACD,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,QAAQ,EAAE,QAAQ,CAAC;IACnB,SAAS,EAAE,CAAC,MAAM,GAAG,SAAS,CAAC,EAAE,CAAC;IAClC,MAAM,CAAC,EAAE,YAAY,CAAC;IACtB,cAAc,CAAC,EAAE,oBAAoB,CAAC;IACtC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iBAAiB,CAAC,EAAE,MAAM,CAAC;CAC5B,iBA0BA;AAED,wBAAsB,+BAA+B,CAAC,EACpD,KAAK,EACL,WAAW,EACX,MAAM,GACP,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,WAAW,EAAE,UAAU,EAAE,CAAC;IAC1B,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC,kBAAkB,EAAE,CAAC,CA6ChC;AAED,wBAAsB,6BAA6B,CAAC,EAClD,KAAK,EACL,WAAW,EACX,sBAAsB,GACvB,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,WAAW,EAAE,oBAAoB,EAAE,CAAC;IACpC,sBAAsB,CAAC,EAAE,sBAAsB,CAAC;CACjD,iBAoDA;AAED,wBAAsB,oBAAoB,CAAC,EACzC,MAAM,EACN,UAAU,GACX,EAAE;IACD,UAAU,EAAE;QACV,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;QAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;KAC5B,CAAC;IACF,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC;IACV,MAAM,EAAE;QAAE,KAAK,EAAE,OAAO,CAAC;QAAC,YAAY,EAAE,MAAM,CAAC;QAAC,QAAQ,EAAE,MAAM,CAAA;KAAE,CAAC;IACnE,cAAc,EAAE,MAAM,CAAC;CACxB,CAAC,CA8BD;AAED,wBAAsB,gBAAgB,CAAC,EACrC,KAAK,EACL,QAAQ,EACR,WAAW,EACX,MAAM,GACP,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,QAAQ,EAAE,QAAQ,CAAC;IACnB,WAAW,EAAE;QACX,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;QAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;KAC5B,EAAE,CAAC;IACJ,MAAM,CAAC,EAAE,YAAY,CAAC;IACtB,cAAc,CAAC,EAAE,oBAAoB,CAAC;IACtC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iBAAiB,CAAC,EAAE,MAAM,CAAC;CAC5B,GAAG,OAAO,CAAC;IAAE,KAAK,EAAE,OAAO,CAAC;IAAC,YAAY,EAAE,MAAM,CAAC;IAAC,QAAQ,EAAE,MAAM,CAAA;CAAE,EAAE,CAAC,CAkIxE"}
+{"version":3,"file":"utils.d.ts","sourceRoot":"","sources":["../../../src/agent/codegen/utils.ts"],"names":[],"mappings":"AAAA,OAAO,EAAkB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AAChE,OAAO,KAAK,EAAE,QAAQ,EAAE,MAAM,4BAA4B,CAAC;AAI3D,OAAO,EAAE,YAAY,EAAE,MAAM,kBAAkB,CAAC;AAMhD,OAAO,EAAE,sBAAsB,EAAE,MAAM,gBAAgB,CAAC;AAExD,OAAO,EAAE,UAAU,EAAE,oBAAoB,EAAE,kBAAkB,EAAE,MAAM,SAAS,CAAC;AAE/E;;;;;;;;;;;GAWG;AACH,wBAAgB,kBAAkB,CAAC,KAAK,EAAE,MAAM,GAAG,UAAU,EAAE,CAiB9D;AAED;;;;;;;;;;GAUG;AACH,wBAAgB,wBAAwB,CACtC,KAAK,EAAE,MAAM,GACZ,oBAAoB,EAAE,CAgBxB;AAED;;;;;;;;;;;;;GAaG;AACH,wBAAgB,wBAAwB,CAAC,KAAK,EAAE,MAAM,GAAG;IACvD,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,cAAc,EAAE,MAAM,GAAG,SAAS,CAAC;IACnC,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;IAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;CAC5B,EAAE,CA8BF;AAED,wBAAgB,2BAA2B,CAAC,KAAK,EAAE,MAAM,GAAG;IAC1D,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,CAAC;IACjB,YAAY,EAAE,MAAM,CAAC;IACrB,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;CACpB,EAAE,CAqCF;AAED,wBAAsB,0BAA0B,CAAC,EAC/C,aAAoB,EACpB,KAAK,EACL,QAAQ,EACR,SAAS,EACT,MAAM,EACN,SAAS,EACT,iBAAiB,GAClB,EAAE;IACD,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,QAAQ,EAAE,QAAQ,CAAC;IACnB,SAAS,EAAE,CAAC,MAAM,GAAG,SAAS,CAAC,EAAE,CAAC;IAClC,MAAM,CAAC,EAAE,YAAY,CAAC;IACtB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iBAAiB,CAAC,EAAE,MAAM,CAAC;CAC5B,iBAyBA;AAED,wBAAsB,+BAA+B,CAAC,EACpD,KAAK,EACL,WAAW,EACX,MAAM,GACP,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,WAAW,EAAE,UAAU,EAAE,CAAC;IAC1B,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC,kBAAkB,EAAE,CAAC,CA6ChC;AAED,wBAAsB,6BAA6B,CAAC,EAClD,KAAK,EACL,WAAW,EACX,sBAAsB,GACvB,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,WAAW,EAAE,oBAAoB,EAAE,CAAC;IACpC,sBAAsB,CAAC,EAAE,sBAAsB,CAAC;CACjD,iBAoDA;AAED,wBAAsB,oBAAoB,CAAC,EACzC,MAAM,EACN,UAAU,GACX,EAAE;IACD,UAAU,EAAE;QACV,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;QAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;KAC5B,CAAC;IACF,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC;IACV,MAAM,EAAE;QAAE,KAAK,EAAE,OAAO,CAAC;QAAC,YAAY,EAAE,MAAM,CAAC;QAAC,QAAQ,EAAE,MAAM,CAAA;KAAE,CAAC;IACnE,cAAc,EAAE,MAAM,CAAC;CACxB,CAAC,CA8BD;AAED,wBAAsB,gBAAgB,CAAC,EACrC,KAAK,EACL,QAAQ,EACR,WAAW,EACX,MAAM,GACP,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,QAAQ,EAAE,QAAQ,CAAC;IACnB,WAAW,EAAE;QACX,QAAQ,EAAE,MAAM,GAAG,SAAS,CAAC;QAC7B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,OAAO,EAAE,MAAM,GAAG,SAAS,CAAC;QAC5B,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC;KAC5B,EAAE,CAAC;IACJ,MAAM,CAAC,EAAE,YAAY,CAAC;IACtB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iBAAiB,CAAC,EAAE,MAAM,CAAC;CAC5B,GAAG,OAAO,CAAC;IAAE,KAAK,EAAE,OAAO,CAAC;IAAC,YAAY,EAAE,MAAM,CAAC;IAAC,QAAQ,EAAE,MAAM,CAAA;CAAE,EAAE,CAAC,CAkIxE"}

package/dist/agent/codegen/utils.js

@@ -127,7 +127,7 @@ function extractTestStepsSuggestions(input) {
     return result.filter((r) => !!r.filePath && !!r.usageExample);
 }
 exports.extractTestStepsSuggestions = extractTestStepsSuggestions;
-async function validateTypesAndFormatCode({ validateTypes = true, trace, testCase, filePaths, logger, testGenOptions, pomPrompt, nonSpecFilePrompt, }) {
+async function validateTypesAndFormatCode({ validateTypes = true, trace, testCase, filePaths, logger, pomPrompt, nonSpecFilePrompt, }) {
     for (let filePath of filePaths) {
         if (!filePath) {
             continue;
@@ -141,7 +141,6 @@ async function validateTypesAndFormatCode({ validateTypes = true, trace, testCas
                     pomCode: pomPrompt ?? "",
                     nonSpecFileCode: nonSpecFilePrompt ?? "",
                     testCase: testCase,
-                    options: testGenOptions,
                 });
             }
             trace?.event({ name: "format-file" });

package/dist/agent/diagnosis-agent/index.d.ts

@@ -1,14 +1,7 @@
 import { TraceClient } from "@empiricalrun/llm";
-import { TestErrorDiagnosisDetails, TestGenConfigOptions } from "@empiricalrun/shared-types";
+import { TestErrorDiagnosisDetails } from "@empiricalrun/shared-types";
 import { CustomLogger } from "../../bin/logger";
-/**
- *
- * inputs
- * - task
- * - diagnosis
- */
-export declare function createTaskUsingFailureDiagnosis({ options, trace, diagnosis, logger, }: {
-    options?: TestGenConfigOptions;
+export declare function createTaskUsingFailureDiagnosis({ trace, diagnosis, logger, }: {
     trace?: TraceClient;
     diagnosis: TestErrorDiagnosisDetails;
     logger?: CustomLogger;

package/dist/agent/diagnosis-agent/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/diagnosis-agent/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAyB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AACvE,OAAO,EACL,yBAAyB,EACzB,oBAAoB,EACrB,MAAM,4BAA4B,CAAC;AAEpC,OAAO,EAAE,YAAY,EAAE,MAAM,kBAAkB,CAAC;AAuChD;;;;;GAKG;AACH,wBAAsB,+BAA+B,CAAC,EACpD,OAAO,EACP,KAAK,EACL,SAAS,EACT,MAAM,GACP,EAAE;IACD,OAAO,CAAC,EAAE,oBAAoB,CAAC;IAC/B,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,SAAS,EAAE,yBAAyB,CAAC;IACrC,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC;IAAE,IAAI,EAAE,MAAM,CAAA;CAAE,CAAC,CA8E5B"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/diagnosis-agent/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAyB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AACvE,OAAO,EAAE,yBAAyB,EAAE,MAAM,4BAA4B,CAAC;AAEvE,OAAO,EAAE,YAAY,EAAE,MAAM,kBAAkB,CAAC;AAuChD,wBAAsB,+BAA+B,CAAC,EACpD,KAAK,EACL,SAAS,EACT,MAAM,GACP,EAAE;IACD,KAAK,CAAC,EAAE,WAAW,CAAC;IACpB,SAAS,EAAE,yBAAyB,CAAC;IACrC,MAAM,CAAC,EAAE,YAAY,CAAC;CACvB,GAAG,OAAO,CAAC;IAAE,IAAI,EAAE,MAAM,CAAA;CAAE,CAAC,CA6E5B"}

package/dist/agent/diagnosis-agent/index.js

@@ -30,13 +30,7 @@ const responseFormat = {
         },
     },
 };
-/**
- *
- * inputs
- * - task
- * - diagnosis
- */
-async function createTaskUsingFailureDiagnosis({ options, trace, diagnosis, logger, }) {
+async function createTaskUsingFailureDiagnosis({ trace, diagnosis, logger, }) {
     trace =
         trace ||
             llm_1.langfuseInstance?.trace({
@@ -48,7 +42,6 @@ async function createTaskUsingFailureDiagnosis({ options, trace, diagnosis, logg
         name: "auto-fix",
         input: {
             diagnosisId: diagnosis.diagnosisId,
-            prjRepoName: options?.metadata.projectRepoName,
         },
     });
     logger?.log("Trying to fix the test using failure diagnosis. Fetching key moments of the diagnosis");

package/dist/agent/enrich-prompt/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/enrich-prompt/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAsB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AAmCpE,eAAO,MAAM,2BAA2B;eAM3B,MAAM;kBACH,MAAM;sBACF,MAAM;;;YAGhB,MAAM;8BACY,OAAO;uBACd,MAAM;EA0D1B,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/enrich-prompt/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAsB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AAmCpE,eAAO,MAAM,2BAA2B;eAM3B,MAAM;kBACH,MAAM;sBACF,MAAM;;;YAGhB,MAAM;8BACY,OAAO;uBACd,MAAM;EAyD1B,CAAC"}

package/dist/agent/enrich-prompt/index.js

@@ -59,7 +59,6 @@ const enrichPromptWithFailingLine = async ({ trace, testBlock, testFilePath, sug
         });
         const llmResponse = await llm.createChatCompletion({
             messages: instructions,
-            trace,
             modelParameters: {
                 temperature: 0.5,
             },

package/dist/agent/infer-agent/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/infer-agent/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAIL,WAAW,EACZ,MAAM,mBAAmB,CAAC;AAC3B,OAAO,EAAE,KAAK,EAAE,oBAAoB,EAAE,MAAM,4BAA4B,CAAC;AAYzE,wBAAsB,mBAAmB,CAAC,EACxC,IAAI,EACJ,OAAO,EACP,KAAK,GACN,EAAE;IACD,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,CAAC,EAAE,oBAAoB,CAAC;IAC/B,KAAK,CAAC,EAAE,WAAW,CAAC;CACrB,GAAG,OAAO,CAAC;IAAE,QAAQ,EAAE,KAAK,CAAA;CAAE,CAAC,CAmE/B"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/agent/infer-agent/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAsB,WAAW,EAAE,MAAM,mBAAmB,CAAC;AACpE,OAAO,EAAE,KAAK,EAAE,oBAAoB,EAAE,MAAM,4BAA4B,CAAC;AAUzE,wBAAsB,mBAAmB,CAAC,EACxC,IAAI,EACJ,OAAO,EACP,KAAK,GACN,EAAE;IACD,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,CAAC,EAAE,oBAAoB,CAAC;IAC/B,KAAK,CAAC,EAAE,WAAW,CAAC;CACrB,GAAG,OAAO,CAAC;IAAE,QAAQ,EAAE,KAAK,CAAA;CAAE,CAAC,CA4D/B"}

package/dist/agent/infer-agent/index.js

@@ -4,16 +4,7 @@ exports.inferAgentBasedTask = void 0;
 const llm_1 = require("@empiricalrun/llm");
 const constants_1 = require("../../constants");
 const promptTemplate_0 = "{{#section \"system\"}}\nYou are a software test engineer specializing in Playwright end-to-end tests. You are given a task which is a part of an\nend-to-end test scenario. The task may involve updating an existing end-to-end test case or writing a new test case from\nscratch. Tests involve user interactions (e.g. click on element) or other actions supported by Playwright (e.g.\nintercept network requests)\n\nYour objective is to identify whether the task requires accessing a web browser or not.\n\nTo fulfill your objective, answer the following questions:\n\n1. Does it require you to interact with a UI element in the browser? Examples of interactions are click, fill, type, key\npress, assert visibility of the element. Actions that interact with network requests are not UI element interactions.\n\n2. Is the locator of this UI element given to you in the task? Locators look like `getByText(...)`, `getByTestId(...)`\nand other locator methods in Playwright\n\n3. Decide if you need a browser: if you need to interact with a UI element AND you are NOT given the locator for that\nelement, you WILL NEED a browser.\n\n4. If you NEED a browser, then respond with answer as \"master\", otherwise respond with \"code\"\n\n\n# Example 1\n## Input\nTask:\nin this test don't delete the agent and remove steps after that\n\n## Output\n- ui_interaction_to_be_performed: There is no interaction here\n- ui_element_to_interact_with: No element specified\n- has_locator_for_that_element: No element specified\n- reasoning_for_browser_required: No interaction hence browser is not required\n- answer: code\n\n# Example 2\n## Input\nTask:\nin the swapfast test, replace the selectTokenForSwap method. Instead we will do this\\nclick on token button - this will\nshow a drawer\\nEnter usd in the search field that shows up in the drawer\\nSelect USDC.axl for Cosmos Hub - very\nimportant to choose this instead of USDC.axl on Osmosis\n\n## Output\n- ui_interaction_to_be_performed: Click on token button\n- ui_element_to_interact_with: Token button\n- has_locator_for_that_element: false\n- reasoning_for_browser_required: Task requires interacting with a UI element and identifying its locator which needs a\nbrowser\n- answer: master\n{{/section}}\n\n{{#section \"user\"}}\nTask:\n{{task}}\n{{/section}}";
-const session_1 = require("../../session");
-const session = (0, session_1.getSessionDetails)();
 async function inferAgentBasedTask({ task, options, trace, }) {
-    trace =
-        trace ||
-            llm_1.langfuseInstance?.trace({
-                name: "infer-agent-task",
-                id: crypto.randomUUID(),
-                release: session.version,
-            });
     const inferAgentSpan = trace?.span({
         name: "infer-agent",
         input: {

package/dist/bin/index.js

@@ -9,6 +9,7 @@ const commander_1 = require("commander");
 const dotenv_1 = __importDefault(require("dotenv"));
 const run_1 = require("../agent/browsing/run");
 const utils_1 = require("../agent/browsing/utils");
+const chat_1 = require("../agent/chat");
 const repo_edit_1 = require("../agent/codegen/repo-edit");
 const run_2 = require("../agent/codegen/run");
 const diagnosis_agent_1 = require("../agent/diagnosis-agent");
@@ -31,14 +32,12 @@ process.on("beforeExit", async () => await flushEvents());
 process.on("exit", async () => await flushEvents());
 process.on("SIGINT", async () => await flushEvents());
 process.on("SIGTERM", async () => await flushEvents());
-async function resolveAgentUsingTask({ testCase, trace, }) {
-    const { response } = await (0, infer_agent_1.inferAgentBasedTask)({
-        task: testCase.steps.join("\n"),
-        trace,
+async function runChatAgent(prompt) {
+    return await (0, chat_1.chatAgent)({
+        prompt,
     });
-    return response;
 }
-async function runAgent(testGenConfig, testGenToken) {
+async function runAgentsWorkflow(testGenConfig, testGenToken) {
     const logger = new logger_1.CustomLogger();
     const { specPath, testCase } = testGenConfig;
     if (process.env.LOG_URL) {
@@ -89,29 +88,10 @@ async function runAgent(testGenConfig, testGenToken) {
         });
         return;
     }
-    // TODO: this needs to be moved to an orchestrator which decides what needs to be done first before executing the sub tasks
-    if (testGenConfig.testErrorDiagnosis &&
-        testGenConfig.testErrorDiagnosis.failingLine &&
-        // TODO: fix this hardcoding of user prompt - ideally its an auto fix intent
-        testCase.steps[0]?.toLowerCase().trim() == "can you please fix the test") {
-        const { task: updatedTask } = await (0, diagnosis_agent_1.createTaskUsingFailureDiagnosis)({
-            options: testGenConfig.options,
-            trace,
-            diagnosis: testGenConfig.testErrorDiagnosis,
-        });
-        if (updatedTask) {
-            testCase.steps = [updatedTask];
-        }
-    }
-    if (!agent || agent === "auto") {
-        agent = await resolveAgentUsingTask({
-            testCase,
-            trace,
-        });
-    }
-    logger.success(`Generating test using ${agent} agent. ${process.env.LOG_URL ? `[view log](${process.env.LOG_URL})` : ""}`);
     if (testGenConfig.testErrorDiagnosis &&
         testGenConfig.testErrorDiagnosis.failingLine) {
+        // If failure context is available, we can enrich the user prompt to contain
+        // the failure context
         const requestedChangeResp = await (0, enrich_prompt_1.enrichPromptWithFailingLine)({
             trace,
             testBlock: testGenConfig.testErrorDiagnosis.failingLine,
@@ -119,7 +99,28 @@ async function runAgent(testGenConfig, testGenToken) {
             suggestionForFix: testCase.steps.join("\n"),
         });
         testCase.steps = [requestedChangeResp.output];
+        // For "auto-fix" we use the user prompt (which is hard-coded in the dashboard
+        // entrypoints), and invoke more interesting enrichment
+        if (
+        // TODO: fix this hardcoding of user prompt - ideally its an auto fix intent
+        testCase.steps[0]?.toLowerCase().trim() == "can you please fix the test") {
+            const { task: updatedTask } = await (0, diagnosis_agent_1.createTaskUsingFailureDiagnosis)({
+                trace,
+                diagnosis: testGenConfig.testErrorDiagnosis,
+            });
+            if (updatedTask) {
+                testCase.steps = [updatedTask];
+            }
+        }
+    }
+    if (!agent || agent === "auto") {
+        const { response } = await (0, infer_agent_1.inferAgentBasedTask)({
+            task: testCase.steps.join("\n"),
+            trace,
+        });
+        agent = response;
     }
+    logger.success(`Generating test using ${agent} agent. ${process.env.LOG_URL ? `[view log](${process.env.LOG_URL})` : ""}`);
     if (agent === "plan") {
         const task = testCase.steps.join("\n");
         const plan = await (0, run_3.planTask)({
@@ -132,11 +133,18 @@ async function runAgent(testGenConfig, testGenToken) {
         await new reporter_1.TestGenUpdatesReporter().sendMessage(plan);
     }
     else if (agent === "code") {
-        await (0, run_2.generateTest)(testCase, specPath, testGenConfig.options, trace);
+        await (0, run_2.generateTestWithCodegen)({
+            testCase,
+            file: specPath,
+            trace,
+        });
     }
     else {
-        // this assumes we have only one scenario in test config
-        const filePathToUpdate = await (0, utils_1.prepareFileForMasterAgent)(testGenConfig, trace);
+        const filePathToUpdate = await (0, utils_1.prepareFileForMasterAgent)({
+            testCase,
+            specPath,
+            trace,
+        });
         void (0, session_1.updateSessionStatus)(testGenConfig.options?.metadata.testSessionId, {
             status: "agent_live_session_started",
         });
@@ -155,10 +163,11 @@ async function runAgent(testGenConfig, testGenToken) {
     const program = new commander_1.Command();
     program
         .option("--token <token>", "Test generation token")
+        .option("--prompt <prompt>", "Prompt for the chat agent")
         .option("--name <test-name>", "Name of the test case")
-        .option("--prompt <prompt>", "Prompt for the test case")
         .option("--file <test-file>", "File path of the test case (inside tests dir)")
         .option("--suites <suites>", "Comma separated list of describe blocks")
+        .option("--use-chat", "Use chat agent (and not the workflow)")
         .parse(process.argv);
     const options = program.opts();
     const completedOptions = await (0, utils_2.validateAndCompleteCliOptions)(options);
@@ -181,10 +190,16 @@ async function runAgent(testGenConfig, testGenToken) {
     });
     let testGenFailed = false;
     let agentUsed;
+    // Download the build if repo has a download script
+    await (0, test_build_1.downloadBuild)(testGenConfig.build || {});
     try {
-        // download the build if it exists
-        await (0, test_build_1.downloadBuild)(testGenConfig.build || {});
-        agentUsed = await runAgent(testGenConfig, testGenToken);
+        if (completedOptions.useChat) {
+            await runChatAgent(completedOptions.prompt);
+            return;
+        }
+        else {
+            agentUsed = await runAgentsWorkflow(testGenConfig, testGenToken);
+        }
     }
     catch (e) {
         testGenFailed = true;

package/dist/bin/utils/index.d.ts

@@ -4,6 +4,7 @@ export interface CliOptions {
     file?: string;
     prompt?: string;
     suites?: string;
+    useChat?: boolean;
 }
 export declare function validateAndCompleteCliOptions(options: CliOptions): Promise<CliOptions>;
 //# sourceMappingURL=index.d.ts.map

package/dist/bin/utils/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/bin/utils/index.ts"],"names":[],"mappings":"AAEA,MAAM,WAAW,UAAU;IACzB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAQD,wBAAsB,6BAA6B,CACjD,OAAO,EAAE,UAAU,GAClB,OAAO,CAAC,UAAU,CAAC,CAmDrB"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/bin/utils/index.ts"],"names":[],"mappings":"AAEA,MAAM,WAAW,UAAU;IACzB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,OAAO,CAAC;CACnB;AAQD,wBAAsB,6BAA6B,CACjD,OAAO,EAAE,UAAU,GAClB,OAAO,CAAC,UAAU,CAAC,CAyDrB"}

package/dist/bin/utils/index.js

@@ -6,12 +6,18 @@ Object.defineProperty(exports, "__esModule", { value: true });
 exports.validateAndCompleteCliOptions = void 0;
 const inquirer_1 = __importDefault(require("inquirer"));
 async function validateAndCompleteCliOptions(options) {
+    // For existing flow between dashboard <> test-gen (via ci-worker)
     const hasToken = !!options.token;
     if (hasToken) {
         return options;
     }
+    let requiredFields = ["name", "file", "prompt"];
+    // For new chat flow in local CLI usage, only prompt is required
+    if (options.useChat) {
+        requiredFields = ["prompt"];
+    }
     const questions = [];
-    if (!options.name) {
+    if (!options.name && requiredFields.includes("name")) {
         questions.push({
             type: "input",
             name: "name",
@@ -19,7 +25,7 @@ async function validateAndCompleteCliOptions(options) {
             validate: (input) => input.trim().length > 0 || "Test name is required",
         });
     }
-    if (!options.file) {
+    if (!options.file && requiredFields.includes("file")) {
         questions.push({
             type: "input",
             name: "file",
@@ -27,7 +33,7 @@ async function validateAndCompleteCliOptions(options) {
             validate: (input) => input.trim().length > 0 || "Test file path is required",
         });
     }
-    if (!options.prompt) {
+    if (!options.prompt && requiredFields.includes("prompt")) {
         questions.push({
             type: "editor",
             name: "prompt",

package/dist/file/server.d.ts

@@ -2,12 +2,14 @@ export declare class FileService {
     private port;
     private filePath;
     private repoDir;
+    private server;
     constructor({ port, repoDir }: {
         port: number;
         repoDir: string;
     });
     setFilePath(filePath: string): void;
     startFileService(): Promise<number>;
+    stop(): Promise<void>;
 }
 export declare function startFileService(): Promise<void>;
 //# sourceMappingURL=server.d.ts.map

package/dist/file/server.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"server.d.ts","sourceRoot":"","sources":["../../src/file/server.ts"],"names":[],"mappings":"AAWA,qBAAa,WAAW;IACtB,OAAO,CAAC,IAAI,CAAa;IACzB,OAAO,CAAC,QAAQ,CAAc;IAC9B,OAAO,CAAC,OAAO,CAAc;gBAEjB,EAAE,IAAI,EAAE,OAAO,EAAE,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,OAAO,EAAE,MAAM,CAAA;KAAE;IAKhE,WAAW,CAAC,QAAQ,EAAE,MAAM;IAItB,gBAAgB,IAAI,OAAO,CAAC,MAAM,CAAC;CAyC1C;AAED,wBAAsB,gBAAgB,kBAAK"}
+{"version":3,"file":"server.d.ts","sourceRoot":"","sources":["../../src/file/server.ts"],"names":[],"mappings":"AAWA,qBAAa,WAAW;IACtB,OAAO,CAAC,IAAI,CAAa;IACzB,OAAO,CAAC,QAAQ,CAAc;IAC9B,OAAO,CAAC,OAAO,CAAc;IAC7B,OAAO,CAAC,MAAM,CAA4C;gBAE9C,EAAE,IAAI,EAAE,OAAO,EAAE,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,OAAO,EAAE,MAAM,CAAA;KAAE;IAKhE,WAAW,CAAC,QAAQ,EAAE,MAAM;IAItB,gBAAgB,IAAI,OAAO,CAAC,MAAM,CAAC;IA0CnC,IAAI,IAAI,OAAO,CAAC,IAAI,CAAC;CAgB5B;AAED,wBAAsB,gBAAgB,kBAAK"}

package/dist/file/server.js

@@ -13,6 +13,7 @@ class FileService {
     port = 0;
     filePath = "";
     repoDir = "";
+    server;
     constructor({ port, repoDir }) {
         this.port = port;
         this.repoDir = repoDir;
@@ -43,7 +44,23 @@ class FileService {
             return res.send({ success: false });
         });
         return new Promise((resolve) => {
-            app.listen(this.port, () => resolve(this.port));
+            this.server = app.listen(this.port, () => resolve(this.port));
+        });
+    }
+    async stop() {
+        return new Promise((resolve, reject) => {
+            if (!this.server) {
+                resolve();
+                return;
+            }
+            this.server.close((err) => {
+                if (err) {
+                    reject(err);
+                    return;
+                }
+                this.server = undefined;
+                resolve();
+            });
         });
     }
 }

package/dist/tools/browser-agent.d.ts

@@ -0,0 +1,16 @@
+import { OpenAI } from "openai";
+export declare const schema: OpenAI.Chat.Completions.ChatCompletionTool;
+export declare const generateTestWithBrowserAgentTool: ({ testName, fileName, changeToMake, }: {
+    testName: string;
+    fileName: string;
+    changeToMake: string;
+}) => Promise<{
+    result: string;
+    gitPatch: string;
+    error?: undefined;
+} | {
+    result: string;
+    error: string;
+    gitPatch?: undefined;
+}>;
+//# sourceMappingURL=browser-agent.d.ts.map

package/dist/tools/browser-agent.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"browser-agent.d.ts","sourceRoot":"","sources":["../../src/tools/browser-agent.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAQhC,eAAO,MAAM,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,WAAW,CAAC,kBA0B5C,CAAC;AAEF,eAAO,MAAM,gCAAgC;cAKjC,MAAM;cACN,MAAM;kBACF,MAAM;;;;;;;;;EAsCrB,CAAC"}