gpt-driver-node 1.0.11 → 1.0.12

package/CHANGELOG.md

@@ -0,0 +1,65 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+See [docs.mobileboost.io](https://docs.mobileboost.io) for full documentation.
+
+## [1.0.12] - 2026-02-27
+
+### Added
+- `maxWaitForStableScreenSecs` config option to control the maximum time to wait for a stable screen
+
+### Changed
+- Enhanced screen stability handling for more reliable test execution
+
+## [1.0.11] - 2026-02-25
+
+### Changed
+- Improved AI execution logging with richer detail (screenshot, prompt, commands, reasoning, cache status)
+
+## [1.0.10] - 2026-02-25
+
+### Added
+- Screen resolution is now automatically sent to the server during session initialization when a device `size` is provided
+
+## [1.0.9] - 2026-02-20
+
+### Changed
+- Smart loop now throws an error on failure instead of resolving silently — update your error handling accordingly
+
+## [1.0.8] - 2026-02-19
+
+### Changed
+- Screenshots are now compressed before being sent, reducing bandwidth usage
+- ADB-based screenshots are used automatically on Android when the Appium server is local
+
+## [1.0.7] - 2026-02-18
+
+### Added
+- `additionalUserContext` option in `GptDriverConfig` — pass extra context to the AI for more accurate test execution
+
+### Fixed
+- Session is now correctly marked as failed when `maxCacheAttempts` is reached
+
+## [1.0.6] - 2026-02-09
+
+### Added
+- Automatic retry logic for middleware service calls
+- ADB screenshot support for Android devices
+- `usemiddleLayerAssertFn` option for assertion handling
+
+### Changed
+- Screen stability checks are now performed once per loop iteration for better cache efficiency
+- Default log level changed to `info`
+
+## [1.0.5] - 2026-01-12
+
+### Added
+- Press/Enter key command support
+- Config file and environment variable support via `configFilePath`
+- `actionDescription` field on commands for improved traceability
+- Screen stability wait between steps to increase cache hit rate
+
+## [1.0.3] - 2026-01-02
+
+### Changed
+- `executeFlow` reliability and performance improvements

package/dist/index.cjs

@@ -632,7 +632,7 @@ async function executeSmartLoop(ctx, params) {
       let commands = [];
       let isCacheHit = false;
       let screenshotResolution = void 0;
-      const stabilityResult = await waitForStableScreen(ctx.getScreenshot);
+      const stabilityResult = await waitForStableScreen(ctx.getScreenshot, { ...ctx.maxWaitForStableScreenSecs && { maxTimeoutSec: ctx.maxWaitForStableScreenSecs } });
       for (let attempt = 0; attempt < maxCacheAttempts; attempt++) {
         const screenshotStartTime = performance.now();
         const screenshotEndTime = performance.now();
@@ -964,6 +964,7 @@ class GptDriver {
   organisationId;
   configFilePath;
   additionalUserContext;
+  maxWaitForStableScreenSecs;
   // Session Execution Stats
   _stats_startTime = 0;
   _stats_executedSteps = 0;
@@ -1000,6 +1001,7 @@ class GptDriver {
     this.organisationId = config.organisationId;
     this.configFilePath = config.configFilePath;
     this.additionalUserContext = config.additionalUserContext;
+    this.maxWaitForStableScreenSecs = config.maxWaitForStableScreenSecs;
     if (config.useGptDriverCloud) {
       if (config.serverConfig.device?.platform == null) {
         throw new Error("Platform is missing. Please specify the platform when using GPTDriver Cloud.");
@@ -1251,7 +1253,8 @@ ${"=".repeat(50)}`);
       logAIExecution: async (params) => this.logAIExecution(params),
       organisationId: this.organisationId,
       middleLayerAssertFn: options?.middleLayerAssertFn,
-      pendingLogPromises: options?.pendingLogPromises
+      pendingLogPromises: options?.pendingLogPromises,
+      maxWaitForStableScreenSecs: this.maxWaitForStableScreenSecs
     };
   }
   /**
@@ -1327,7 +1330,7 @@ ${"=".repeat(50)}`);
         actions: [
           { type: "pointerMove", duration: 0, x: clampedX, y: clampedY },
           { type: "pointerDown", button: 0 },
-          { type: "pause", duration: 100 },
+          { type: "pause", duration: 300 },
           { type: "pointerUp", button: 0 }
         ]
       }
@@ -1580,7 +1583,7 @@ ${"=".repeat(50)}`);
     const driver = this.driver;
     if (appiumHandler != null) {
       try {
-        await this.takeScreenshotAndLogCodeExecution(appiumHandler.toString());
+        this.pendingLogPromises.push(this.takeScreenshotAndLogCodeExecution(appiumHandler.toString()));
         await appiumHandler(driver);
         globalLogger.debug("Custom Appium handler executed successfully");
       } catch (e) {
@@ -1612,7 +1615,7 @@ ${"=".repeat(50)}`);
       try {
         const handlerStartTime = performance.now();
         globalLogger.debug(`[Performance] Executing custom Appium handler...`);
-        await this.takeScreenshotAndLogCodeExecution(appiumHandler.toString());
+        this.pendingLogPromises.push(this.takeScreenshotAndLogCodeExecution(appiumHandler.toString()));
         await appiumHandler(driver);
         const handlerEndTime = performance.now();
         globalLogger.debug("Custom Appium handler executed successfully");

package/dist/index.d.cts

@@ -25,6 +25,7 @@ interface GptDriverConfig {
     additionalUserContext?: string;
     organisationId?: string;
     configFilePath?: string;
+    maxWaitForStableScreenSecs?: number;
 }
 /**
  * Parameters for opening a deep link url in the Appium session.
@@ -514,6 +515,7 @@ declare class GptDriver {
     private organisationId?;
     private configFilePath?;
     private additionalUserContext?;
+    private maxWaitForStableScreenSecs?;
     private _stats_startTime;
     private _stats_executedSteps;
     private _stats_cacheHits;

package/dist/index.mjs

@@ -630,7 +630,7 @@ async function executeSmartLoop(ctx, params) {
       let commands = [];
       let isCacheHit = false;
       let screenshotResolution = void 0;
-      const stabilityResult = await waitForStableScreen(ctx.getScreenshot);
+      const stabilityResult = await waitForStableScreen(ctx.getScreenshot, { ...ctx.maxWaitForStableScreenSecs && { maxTimeoutSec: ctx.maxWaitForStableScreenSecs } });
       for (let attempt = 0; attempt < maxCacheAttempts; attempt++) {
         const screenshotStartTime = performance.now();
         const screenshotEndTime = performance.now();
@@ -962,6 +962,7 @@ class GptDriver {
   organisationId;
   configFilePath;
   additionalUserContext;
+  maxWaitForStableScreenSecs;
   // Session Execution Stats
   _stats_startTime = 0;
   _stats_executedSteps = 0;
@@ -998,6 +999,7 @@ class GptDriver {
     this.organisationId = config.organisationId;
     this.configFilePath = config.configFilePath;
     this.additionalUserContext = config.additionalUserContext;
+    this.maxWaitForStableScreenSecs = config.maxWaitForStableScreenSecs;
     if (config.useGptDriverCloud) {
       if (config.serverConfig.device?.platform == null) {
         throw new Error("Platform is missing. Please specify the platform when using GPTDriver Cloud.");
@@ -1249,7 +1251,8 @@ ${"=".repeat(50)}`);
       logAIExecution: async (params) => this.logAIExecution(params),
       organisationId: this.organisationId,
       middleLayerAssertFn: options?.middleLayerAssertFn,
-      pendingLogPromises: options?.pendingLogPromises
+      pendingLogPromises: options?.pendingLogPromises,
+      maxWaitForStableScreenSecs: this.maxWaitForStableScreenSecs
     };
   }
   /**
@@ -1325,7 +1328,7 @@ ${"=".repeat(50)}`);
         actions: [
           { type: "pointerMove", duration: 0, x: clampedX, y: clampedY },
           { type: "pointerDown", button: 0 },
-          { type: "pause", duration: 100 },
+          { type: "pause", duration: 300 },
           { type: "pointerUp", button: 0 }
         ]
       }
@@ -1578,7 +1581,7 @@ ${"=".repeat(50)}`);
     const driver = this.driver;
     if (appiumHandler != null) {
       try {
-        await this.takeScreenshotAndLogCodeExecution(appiumHandler.toString());
+        this.pendingLogPromises.push(this.takeScreenshotAndLogCodeExecution(appiumHandler.toString()));
         await appiumHandler(driver);
         globalLogger.debug("Custom Appium handler executed successfully");
       } catch (e) {
@@ -1610,7 +1613,7 @@ ${"=".repeat(50)}`);
       try {
         const handlerStartTime = performance.now();
         globalLogger.debug(`[Performance] Executing custom Appium handler...`);
-        await this.takeScreenshotAndLogCodeExecution(appiumHandler.toString());
+        this.pendingLogPromises.push(this.takeScreenshotAndLogCodeExecution(appiumHandler.toString()));
         await appiumHandler(driver);
         const handlerEndTime = performance.now();
         globalLogger.debug("Custom Appium handler executed successfully");

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "gpt-driver-node",
-  "version": "1.0.11",
+  "version": "1.0.12",
   "main": "./dist/index.cjs",
   "module": "./dist/index.mjs",
   "types": "./dist/index.d.cts",