testchimp-runner-core 0.0.25 → 0.0.28

package/src/scenario-worker-class.ts

@@ -97,8 +97,6 @@ export class ScenarioWorker extends EventEmitter {
    * Initialize orchestrator mode with tools
    */
   private initializeOrchestrator(): void {
-    this.log('🤖 Initializing Orchestrator Mode');
-    
     // Create tool registry
     this.toolRegistry = new ToolRegistry();
     
@@ -135,12 +133,12 @@ export class ScenarioWorker extends EventEmitter {
       this.debugMode  // Pass debug mode
     );
     
-    this.log(`✓ Orchestrator initialized with 5 tools${this.debugMode ? ' (DEBUG MODE)' : ''} (information-gathering only)`);
+    // Minimal initialization logging - internal details not needed by consumer
   }
 
   private log(message: string): void {
-    const timestamp = new Date().toISOString().substring(11, 23); // HH:MM:SS.mmm
-    const formattedMessage = `[${timestamp}] [ScenarioWorker] ${message}`;
+    // Let consumer add timestamps - just report the raw message
+    const formattedMessage = `[ScenarioWorker] ${message}`;
     // Always log to console for debug visibility
     console.log(formattedMessage);
     // Also route to outputChannel if provided
@@ -150,8 +148,8 @@ export class ScenarioWorker extends EventEmitter {
   }
 
   private logError(message: string): void {
-    const timestamp = new Date().toISOString().substring(11, 23); // HH:MM:SS.mmm
-    const formattedMessage = `[${timestamp}] [ScenarioWorker] ERROR: ${message}`;
+    // Let consumer add timestamps - just report the raw message
+    const formattedMessage = `[ScenarioWorker] ERROR: ${message}`;
     // Always log to console for debug visibility
     console.error(formattedMessage);
     // Also route to outputChannel if provided
@@ -246,21 +244,16 @@ export class ScenarioWorker extends EventEmitter {
 
   async initialize(): Promise<void> {
     try {
-      const RUNNER_CORE_VERSION = "v1.5.0-vision-preserve-values";
-      this.log('═══════════════════════════════════════════════════════');
-      this.log(`🚀 RUNNER-CORE VERSION: ${RUNNER_CORE_VERSION}`);
-      this.log('═══════════════════════════════════════════════════════');
-      this.log('Initializing Scenario worker...');
       this.sessionId = `scenario_worker_${Date.now()}`;
       this.initialized = true;
-      this.log(`Scenario worker initialized with session: ${this.sessionId}`);
+      // Minimal initialization - consumer doesn't need to see internal details
     } catch (error) {
       this.logError(`Scenario worker initialization error: ${error}`);
       throw error;
     }
   }
 
-  async processScenarioJob(job: ScenarioJob): Promise<ScenarioResponse> {
+  async processScenarioJob(job: ScenarioRunJob): Promise<ScenarioResponse> {
     if (!this.initialized) {
       throw new Error('Scenario worker not initialized');
     }
@@ -268,9 +261,9 @@ export class ScenarioWorker extends EventEmitter {
     // Set current job ID for progress reporting
     this.currentJobId = job.id;
 
-    // VERSION MARKER - increment this number with each significant change
-    const RUNNER_CORE_VERSION = "v1.5.0-vision-preserve-values";
-    this.log(`🚀 RUNNER-CORE VERSION: ${RUNNER_CORE_VERSION}`);
+    // Log library version once (read from package.json)
+    const packageJson = require('../package.json');
+    this.log(`testchimp-runner-core v${packageJson.version}`);
     this.log(`📋 Processing scenario: ${job.scenario}`);
 
     const startTime = Date.now();
@@ -310,16 +303,30 @@ export class ScenarioWorker extends EventEmitter {
       }
       this.emit('log', job.id, `\n## Execution Progress\n\n`);
 
-      // 2. Start a new browser session using centralized utility
-      // Default to headed mode (headless: false) for better debugging
-      // Create logger function from outputChannel for browser initialization
-      const logger = this.outputChannel ? (message: string, level?: 'log' | 'error' | 'warn') => {
-        this.outputChannel!.appendLine(`[Browser] ${message}`);
-      } : undefined;
-      const browserInstance = await initializeBrowser(job.playwrightConfig, false, undefined, logger);
-      browser = browserInstance.browser;
-      context = browserInstance.context;
-      page = browserInstance.page;
+      // 2. Start a new browser session or use existing one
+      if (job.existingBrowser && job.existingContext && job.existingPage) {
+        // Use existing browser provided by caller (e.g., scriptservice)
+        this.log('Using existing browser/page provided by caller');
+        browser = job.existingBrowser;
+        context = job.existingContext;
+        page = job.existingPage;
+      } else {
+        // Create new browser (default behavior for local clients)
+        // Default to headed mode (headless: false) for better debugging
+        // Create logger function from outputChannel for browser initialization
+        const logger = this.outputChannel ? (message: string, level?: 'log' | 'error' | 'warn') => {
+          this.outputChannel!.appendLine(`[Browser] ${message}`);
+        } : undefined;
+        const browserInstance = await initializeBrowser(job.playwrightConfig, false, undefined, logger);
+        browser = browserInstance.browser;
+        context = browserInstance.context;
+        page = browserInstance.page;
+      }
+      
+      // LIFECYCLE: Call beforeStartTest if provided
+      if (this.progressReporter?.beforeStartTest) {
+        await this.progressReporter.beforeStartTest(page, browser, context);
+      }
       
         // Set reasonable timeout for most operations
         // 5 seconds for element interactions (fast feedback on wrong selectors)
@@ -367,6 +374,17 @@ export class ScenarioWorker extends EventEmitter {
           step.stepNumber = i + 1;
           
           try {
+            // LIFECYCLE: Call beforeStepStart if provided
+            if (this.progressReporter?.beforeStepStart) {
+              await this.progressReporter.beforeStepStart(
+                {
+                  stepNumber: step.stepNumber,
+                  description: step.description
+                },
+                page
+              );
+            }
+            
             // Use orchestrator to execute this step
             const result = await this.orchestratorAgent.executeStep(
               page,
@@ -405,6 +423,18 @@ export class ScenarioWorker extends EventEmitter {
               }
             }
             
+            // REPORT FINAL STEP RESULT (after orchestrator completes all iterations)
+            // This gives the complete accumulated commands, not just one iteration
+            await this.reportStepProgress({
+              jobId: job.id,
+              stepNumber: step.stepNumber,
+              description: step.description,
+              code: step.playwrightCommands?.join('\n') || '',  // All accumulated commands
+              status: step.success ? StepExecutionStatus.SUCCESS : StepExecutionStatus.FAILURE,
+              error: step.error,
+              agentIteration: result.iterations
+            });
+            
           } catch (error: any) {
             this.logError(`Orchestrator execution failed for step ${step.stepNumber}: ${error.message}`);
             step.success = false;
@@ -858,7 +888,17 @@ export class ScenarioWorker extends EventEmitter {
       }
 
       // Generate clean script with TestChimp comment and code
+      this.log(`[ScenarioWorker] Generating script from ${steps.length} steps`);
+      steps.forEach((s, i) => {
+        this.log(`[ScenarioWorker]   Step ${i+1}: ${s.description}`);
+        this.log(`[ScenarioWorker]     Commands: ${s.playwrightCommands?.length || 0}`);
+        if (s.playwrightCommands && s.playwrightCommands.length > 0) {
+          this.log(`[ScenarioWorker]     First command: ${s.playwrightCommands[0]}`);
+        }
+      });
+      
       generatedScript = generateTestScript(testName, steps, undefined, hashtags);
+      this.log(`[ScenarioWorker] Generated script length: ${generatedScript.length}`);
 
       // Perform final cleanup pass to remove redundancies and make minor adjustments
       this.log(`[ScenarioWorker] Performing final script cleanup...`);
@@ -931,6 +971,15 @@ export class ScenarioWorker extends EventEmitter {
       
       const executionLog = logLines.join('\n');
 
+      // Report job completion
+      await this.reportJobProgress({
+        jobId: job.id,
+        status: overallSuccess ? 'completed' : 'failed',
+        testName,
+        script: generatedScript,
+        error: overallSuccess ? undefined : 'Some steps failed during execution'
+      });
+
       return {
         success: overallSuccess,
         steps,
@@ -944,6 +993,16 @@ export class ScenarioWorker extends EventEmitter {
     } catch (error: any) {
       overallSuccess = false;
       this.logError(`Overall scenario processing error: ${error}`);
+      
+      // Report job failure
+      await this.reportJobProgress({
+        jobId: job.id,
+        status: 'failed',
+        testName: job.testName || 'test',
+        script: generatedScript,
+        error: error instanceof Error ? error.message : 'Unknown error during scenario processing'
+      });
+      
       return {
         success: false,
         steps,
@@ -955,7 +1014,22 @@ export class ScenarioWorker extends EventEmitter {
         error: error instanceof Error ? error.message : 'Unknown error during scenario processing'
       };
     } finally {
-      if (browser) {
+      // LIFECYCLE: Call afterEndTest if provided
+      if (browser && this.progressReporter?.afterEndTest) {
+        try {
+          await this.progressReporter.afterEndTest(
+            overallSuccess ? 'passed' : 'failed',
+            overallSuccess ? undefined : 'Test execution had failures',
+            page
+          );
+        } catch (callbackError) {
+          this.log(`afterEndTest callback failed: ${callbackError}`);
+        }
+      }
+      
+      // Only close browser if we created it (not provided by caller)
+      const usingExternalBrowser = !!(job.existingBrowser && job.existingContext && job.existingPage);
+      if (browser && !usingExternalBrowser) {
         await browser.close();
       }
     }

package/src/types.ts

@@ -83,6 +83,11 @@ export interface ScenarioRunJob {
   playwrightConfig?: PlaywrightConfig;
   model?: string;
   scenarioFileName?: string;
+  
+  // Optional: Provide existing browser/page/context (for server-side usage)
+  existingBrowser?: any;
+  existingContext?: any;
+  existingPage?: any;
 }
 
 /**
@@ -139,6 +144,11 @@ export interface ScenarioJob {
   config?: PlaywrightConfig;
   resolve: (result: ScenarioResponse) => void;
   reject: (error: Error) => void;
+  
+  // Optional: Provide existing browser/page/context (for server-side usage)
+  existingBrowser?: any;
+  existingContext?: any;
+  existingPage?: any;
 }
 
 // ============================================================================
@@ -166,6 +176,12 @@ export interface ScriptExecutionRequest {
   model?: string;
   headless?: boolean; // defaults to false (headed)
   deflake_run_count?: number; // defaults to 1
+  
+  // Optional: Provide existing browser/page/context (for server-side usage)
+  // If not provided, runner-core will create its own
+  existingBrowser?: any; // Browser instance
+  existingContext?: any; // BrowserContext instance
+  existingPage?: any; // Page instance
 }
 
 /**

package/RELEASE_0.0.23.md

@@ -1,120 +0,0 @@
-# Release 0.0.23 - Published to npm
-
-## Summary
-Published `testchimp-runner-core@0.0.23` to npm registry and updated `testchimp-vs-extension` to use the published package instead of local file references.
-
-## Changes in 0.0.23
-
-### New Features
-1. **Script Cleanup Feature** - Final LLM-based cleanup pass to remove redundancies
-2. **Semantic Selector Preference** - Prioritize getByRole, getByLabel, etc. over auto-generated IDs
-3. **Playwright Expect Assertions** - Use expect() for verifications instead of manual checks
-4. **Focused Step Execution** - Only execute what's explicitly asked in each step
-
-### Improvements
-1. Enhanced selector generation priority in page-info-utils
-2. Improved orchestrator prompts for better code quality
-3. Fixed comment placement in generated scripts
-4. Made expect() available in command execution context
-5. Added comprehensive guidance to avoid extra assertions
-
-### Bug Fixes
-1. Fixed duplicate/orphaned comments in generated scripts
-2. Removed self-dependency from package.json
-3. Fixed comment ordering (now always appears before code)
-
-## Publishing Details
-
-### Version Update
-- **Previous:** 0.0.22 (local tarball)
-- **Current:** 0.0.23 (npm published)
-- **Registry:** https://registry.npmjs.org/
-- **Package Size:** 14.0 MB
-- **Unpacked Size:** 14.8 MB
-- **Files:** 166
-
-### Package Changes
-**runner-core/package.json:**
-- Updated version from 0.0.22 → 0.0.23
-- Removed self-dependency on testchimp-runner-core
-
-**vs-ext/package.json:**
-- Changed from: `"testchimp-runner-core": "file:testchimp-runner-core-0.0.22.tgz"`
-- Changed to: `"testchimp-runner-core": "^0.0.23"`
-
-### Cleanup Performed
-- Removed all local .tgz tarball files
-- Removed testchimp-runner-core@0.0.22 directory from vs-ext
-- Clean installation from npm registry
-
-## Verification
-
-```bash
-$ npm list testchimp-runner-core
-testchimp-vs-extension@0.0.8
-└── testchimp-runner-core@0.0.23
-```
-
-✅ **vs-ext successfully using published npm package**
-
-## Build Status
-
-✅ **runner-core**
-- Clean build successful
-- TypeScript compilation: ✓
-- Published to npm: ✓
-
-✅ **vs-ext**  
-- Dependencies installed from npm: ✓
-- Webpack production build: ✓ (492 KiB)
-- Extension bundle: Ready
-
-## Benefits of npm Publishing
-
-### For Development
-1. **Version Control** - Clear versioning in package.json
-2. **Dependency Management** - Standard npm workflow
-3. **Faster Installs** - No need to rebuild tarball
-4. **Better CI/CD** - Can reference specific versions
-
-### For Distribution
-1. **Public Access** - Anyone can install via npm
-2. **Version History** - All versions tracked on npm
-3. **Automatic Updates** - Use semver ranges (^0.0.23)
-4. **No Local Files** - Cleaner repository
-
-## Next Steps
-
-### For Future Releases
-1. Update version in runner-core/package.json
-2. Build: `npm run build`
-3. Publish: `npm publish`
-4. Update vs-ext dependency to new version
-5. Install and rebuild vs-ext
-
-### Deployment
-The vs-ext can now be:
-- Packaged as .vsix with the npm-installed dependency
-- Distributed without including runner-core tarball
-- Updated independently by bumping the version range
-
-## Documentation
-
-All features are documented in:
-- `plandocs/SELECTOR_IMPROVEMENTS.md` - Selector preference strategy
-- `plandocs/SCRIPT_CLEANUP_FEATURE.md` - Cleanup feature details
-- This file - Release notes
-
-## Breaking Changes
-
-None. All changes are backward compatible.
-
-## Migration Notes
-
-For anyone using testchimp-runner-core@0.0.22:
-1. Update package.json to: `"testchimp-runner-core": "^0.0.23"`
-2. Run: `npm install`
-3. Rebuild your project
-
-No code changes required - all improvements are in the core library.
-

package/RELEASE_0.0.24.md

@@ -1,161 +0,0 @@
-# Release 0.0.24 - Suppress Token Usage Logs in Staging/Production
-
-## Summary
-Token usage logs are now suppressed in staging and production environments to reduce noise in production logs. Detailed token usage information is only shown in development mode (localhost).
-
-## Changes in 0.0.24
-
-### Environment-Aware Logging
-Added intelligent environment detection to suppress verbose logs in production environments:
-
-1. **New Function: `isDevelopmentMode()`** (`env-loader.ts`)
-   - Detects if running in development mode based on:
-     - Backend URL contains `localhost` or `127.0.0.1`
-     - `NODE_ENV` is explicitly set to `'development'`
-   - Returns `false` for staging/production (safe default)
-
-2. **Suppressed Logs in Production**
-   - Token usage logs: `💰 Reporting token usage: X + Y`
-   - Missing usage data warnings: `⚠ No usage data in LLM response`
-   - These logs still appear when running locally for debugging
-
-### Implementation Details
-
-**env-loader.ts:**
-```typescript
-export function isDevelopmentMode(): boolean {
-  try {
-    const config = loadEnvConfig();
-    const backendUrl = config.TESTCHIMP_BACKEND_URL || '';
-    
-    // Development mode if:
-    // 1. Backend URL contains localhost or 127.0.0.1
-    // 2. NODE_ENV is explicitly set to 'development'
-    const isLocalhost = backendUrl.includes('localhost') || backendUrl.includes('127.0.0.1');
-    const isDevEnv = process.env.NODE_ENV === 'development';
-    
-    return isLocalhost || isDevEnv;
-  } catch (error) {
-    // If we can't determine, assume production (suppress logs)
-    return false;
-  }
-}
-```
-
-**orchestrator-agent.ts:**
-```typescript
-// Only log token usage in development mode
-if (isDevelopmentMode()) {
-  this.logger?.(`[Orchestrator] 💰 Reporting token usage: ${tokenUsage.inputTokens} + ${tokenUsage.outputTokens}`, 'log');
-}
-
-// Only log missing usage data in development mode  
-else if (!response.usage && isDevelopmentMode()) {
-  this.logger?.(`[Orchestrator] ⚠ No usage data in LLM response`, 'warn');
-}
-```
-
-## Environment Detection Logic
-
-### Development Mode (logs shown)
-- Backend URL: `http://localhost:*` or `http://127.0.0.1:*`
-- OR `NODE_ENV=development`
-
-### Staging/Production (logs suppressed)
-- env.staging: `https://featureservice-staging.testchimp.io`
-- env.prod: `https://featureservice.testchimp.io`
-
-## Benefits
-
-1. **Cleaner Production Logs** - No verbose token usage logs in staging/prod
-2. **Better User Experience** - Users don't see internal debugging information
-3. **Still Debug-Friendly** - Developers still see all logs when running locally
-4. **Safe Defaults** - If environment can't be determined, assume production (suppress logs)
-5. **Token Tracking Still Works** - `progressReporter.onTokensUsed()` is still called regardless of logging
-
-## What Still Happens in Production
-
-Even though logs are suppressed, the following still works:
-- ✅ Token usage is tracked via `progressReporter.onTokensUsed()`
-- ✅ Token metrics are sent to backend
-- ✅ All functionality remains unchanged
-- ✅ Error logs and warnings (non-token-related) still appear
-
-## What's Suppressed in Production
-
-Only these specific verbose logs are hidden:
-- ❌ `💰 Reporting token usage: X + Y`
-- ❌ `⚠ No usage data in LLM response`
-
-## Testing
-
-### Local Development
-```bash
-# Backend URL will be localhost
-# OR NODE_ENV=development
-# Result: Token usage logs VISIBLE
-```
-
-### Staging
-```bash
-# env.staging: https://featureservice-staging.testchimp.io
-# Result: Token usage logs HIDDEN
-```
-
-### Production  
-```bash
-# env.prod: https://featureservice.testchimp.io
-# Result: Token usage logs HIDDEN
-```
-
-## Migration
-
-No code changes required. Simply update to `^0.0.24`:
-
-```json
-{
-  "dependencies": {
-    "testchimp-runner-core": "^0.0.24"
-  }
-}
-```
-
-## Files Modified
-
-1. `/src/env-loader.ts` - Added `isDevelopmentMode()` function
-2. `/src/orchestrator/orchestrator-agent.ts` - Added environment checks to token usage logs
-
-## Backward Compatibility
-
-✅ Fully backward compatible - no breaking changes
-✅ All functionality preserved
-✅ Only logging behavior changed
-
-## Version History
-
-- **0.0.24** - Environment-aware token usage logging
-- **0.0.23** - Script cleanup feature, semantic selectors, expect() assertions
-- **0.0.22** - Previous version (local tarball)
-
-## Published to npm
-
-```
-✅ Published: testchimp-runner-core@0.0.24
-📦 Package Size: 243.4 kB (significantly smaller!)
-📋 Registry: https://registry.npmjs.org/
-```
-
-## Next Steps
-
-For future releases, any verbose development logs should use:
-
-```typescript
-import { isDevelopmentMode } from '../env-loader';
-
-if (isDevelopmentMode()) {
-  this.logger?.('Detailed debug information', 'log');
-}
-```
-
-This ensures production logs stay clean while maintaining full debugging capabilities for local development.
-