Merge pull request #8511 from andreas-mitrou/add-browser-automation-docs

AB#5041: Add browser automation docs

Author: lucciz01

support/power-platform/power-automate/desktop-flows/browser-automation/failed-to-assume-control-of-edge-chrome-firefox-error.md

@@ -2,106 +2,171 @@
 title: Failed to assume control of Edge or Chrome or Firefox error
 description: Provides a resolution to the error that occurs when running a desktop flow that has a "Launch browser" action in Power Automate.
 ms.reviewer: nimoutzo, jefernn
-ms.date: 05/24/2023
+ms.date: 04/09/2025
 ms.custom: sap:Desktop flows\UI or browser automation
 ---
 # "Failed to assume control of Microsoft Edge/Chrome/Firefox" error
 
-This article provides a resolution for the error that may occur when you run a desktop flow that contains a "Launch browser" action in Microsoft Power Automate. If the error occurs right after a browser update and has the same symptoms, see [Browser automation actions stop working after a browser update](browser-automation-error-after-chromium-update.md).
+This article provides a resolution for the error that might occur when you run a desktop flow that contains a Launch Browser action in Microsoft Power Automate. If the error occurs right after a browser update and has the same symptoms, see [Browser automation actions stop working after a browser update](browser-automation-error-after-chromium-update.md).
 
 _Applies to:_   Power Automate  
 _Original KB number:_   5001691
 
 ## Symptoms
 
-When you run a desktop flow that has a "Launch Edge", "Launch Chrome", or "Launch Firefox" action in Microsoft Power Automate, the execution fails with one of the following error messages:
+### Scenario 1
 
-> Failed to assume control of Microsoft Edge (Internal error or communication failure).
+When you run a desktop flow that has a [Launch new Microsoft Edge](/power-automate/desktop-flows/actions-reference/webautomation#launch-new-microsoft-edge), [Launch new Chrome](/power-automate/desktop-flows/actions-reference/webautomation#launchchromebase), or [Launch new Firefox](/power-automate/desktop-flows/actions-reference/webautomation#launchfirefoxbase) action in Microsoft Power Automate, the execution fails with one of the following error messages:
 
-> Failed to assume control of Chrome (Internal error or communication failure).
+- > Failed to assume control of Microsoft Edge (Internal error or communication failure).
+- > Failed to assume control of Chrome (Internal error or communication failure).
+- > Failed to assume control of Firefox (Internal error or communication failure)
 
-> Failed to assume control of Firefox (Internal error or communication failure)
+> [!IMPORTANT]
+> To resolve the error:
+>
+> 1. First run the **Troubleshoot UI/Web automation issues** diagnostic using the [Power Automate for desktop troubleshooter](/power-automate/desktop-flows/troubleshooter).  
+> 2. When the diagnostic runs, a report is generated that identifies issues. These issues can be resolved by pressing the **Fix** button that appears after the diagnostics check completes.  
+> 3. If the troubleshooter doesn't resolve the error, proceed to the potential causes and resolutions provided in this article.
+
+### Scenario 2
+
+The error also occurs when you run a Launch Browser action, and the browser is launched using a different system user than the one used to run Power Automate for desktop. For the recommended workaround, see [Cause 3](#cause-3-the-browser-is-launched-using-a-different-system-user-than-the-one-used-to-run-power-automate-for-desktop-version-238-or-higher) in this article.
 
 ## Cause 1: Web extension isn't installed properly or enabled
 
 The Microsoft Edge, Google Chrome, or Firefox web extension isn't installed properly or enabled.
 
-## Verifying issue for cause 1
+### Verifying issue for cause 1
 
 - Execution of a desktop flow with one of the respective actions fails with the error message.
-- Web Recorder initiation for the specific browser shows the below message:
+- Web Recorder initiation for the specific browser shows the following message:
 
-  :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/you-need-power-automate-desktop-extension.png" alt-text="A message that contains a Get Extension button to install the web extension.":::
+  :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/you-need-power-automate-desktop-extension.png" alt-text="Screenshot that shows a message containing a Get Extension button to install the web extension.":::
 
-- Note that the web extension might seem to be installed and enabled in the extension repository of the specific browser but the execution to be unsuccessful.
+- The web extension might seem to be installed and enabled in the extension repository of the specific browser, but it fails to execute.
 
 ### Resolution
 
 1. Install the respective web extension from the Power Automate for desktop designer.
 
-   :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/browser-extensions.png" alt-text="Select to install the web extension from the list.":::
+   :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/browser-extensions.png" alt-text="Screenshot that shows how to install the web extension from the list.":::
 
 2. Even if the extension seems installed and enabled, you need to remove it and reinstall it.
 3. Restart the browser.
 
-## Cause 2: Launch browser action takes longer than the default timeout
+## Cause 2: Launch Browser action takes longer than the default timeout
 
-Launch of the browser takes longer than the default timeout (30 seconds) of the action. The automation tries to assume control of the browser before opening and as a result the "Launch browser" action fails.
+Launch of the browser takes longer than the default timeout (**30** seconds) of the action. The automation tries to assume control of the browser before opening and as a result the Launch Browser action fails.
 
-## Verifying issue for cause 2
+### Verifying issue for cause 2
 
 Execution of a desktop flow with one of the respective actions fails with the error message. This behavior might not be consistent (some executions could be successful).
 
-### Resolution
+### Resolution 1
+
+In the **Launch new Edge**, **Launch new Chrome**, or **Launch new Firefox** action, increase the values of **Timeout on webpage load** and **Timeout** parameters located in the **Advanced** section. For example, set these values to **120** seconds.
+
+If this adjustment doesn't resolve the issue, follow the steps outlined in [Resolution 2](#resolution-2) or [Resolution 3](#resolution-3).
 
-1. Insert a new "Launch new browser" action as:
-    - Launch new Edge
-    - Launch new Chrome
-    - Launch new Firefox
-    - Launch new Internet Explorer
+### Resolution 2
+
+Use the Launch Browser action to start the process of the corresponding browser. Then a combination of an additional Launch Browser action with mode set to **Attach to running instance** and a [Go to web page](/power-automate/desktop-flows/actions-reference/webautomation#gotowebpagebase) action afterwards can be used to resolve the issue.
+
+1. Insert a new Launch Browser action as:
+    - **Launch new Microsoft Edge**
+    - **Launch new Chrome**
+    - **Launch new Firefox**
+    - **Launch new Internet Explorer**
 
 2. Set up the parameters of the action:
     - Launch mode: Launch new Instance
     - Initial URL: A default URL
-    - Rest parameters can be set as desired.
+    - The remaining parameters can be set as desired.
 
-      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/parameters-of-launch-new-browser-action.png" alt-text="Set up the parameters of the Launch new browser action.":::
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/parameters-of-launch-new-browser-action.png" alt-text="Screenshot that shows how to set up the parameters of the Launch new browser action.":::
 
-3. Insert an error handling policy by:
+3. Insert an error handling policy:
     - Select the **On error** option in the action window:
 
-      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/insert-error-handling-policy-for-action.png" alt-text="Insert an error handling policy by using the On error option.":::
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/insert-error-handling-policy-for-action.png" alt-text="Screenshot that shows how to insert an error handling policy by using the On error option.":::
 
     - Select **Continue flow run** > **Go to next action** in the dropdown list, and then select **Save**.
 
-      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/save-settings-on-error.png" alt-text="Select the Continue flow run and Go to next action options and then save the settings.":::
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/save-settings-on-error.png" alt-text="Screenshot that shows how to select the Continue flow run and Go to next action options and then save the settings.":::
 
-4. Insert a new "Launch new browser" action of the same browser as step 1 and set up the parameters of the action:
+4. Insert a new Launch Browser action of the same browser as step 1 and set up the parameters of the action:
     - Launch mode: Attach to running instance
     - Attach to browser tab: By URL
     - Tab URL: The URL inserted in step 1.
-    - Variables produced: Replace the new Browser to variable to the name of the variable produced by step 1.
+    - Variables produced: Replace the new browser variable with the name of the variable produced in step 1.
 
-        :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/replace-variables-produced-browser.png" alt-text="Replace the new Browser to variable to the name of the variable produced by step 1.":::
+        :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/replace-variables-produced-browser.png" alt-text="Screenshot that shows how to replace the new browser variable with the name of the variable produced in step 1.":::
 
 5. Select the **On error** option and take the following steps:
     1. Enable the **Retry action if an error occurs** option.
 
-       :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/enable-retry-action-if-an-error-occurs.png" alt-text="Enable the Retry action if an error occurs option.":::
+       :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/enable-retry-action-if-an-error-occurs.png" alt-text="Screenshot that shows how to enable the Retry action if an error occurs option.":::
 
     2. Set the number of times to **20** by selecting the number of times.
     3. Set the interval in seconds to **5** by selecting the number of seconds.
     4. Select **Save**.
 
-6. Insert a "Go to Web page" action and set up the parameters of the action:
-    - Web browser instance: The variable produced by the "Launch new browser" action.
+6. Insert a **Go to web page** action and set up the parameters of the action:
+    - Web browser instance: The variable produced by the Launch Browser action.
     - Navigate: To URL
     - URL: The URL you would like to navigate to.
 
-      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/select-parameters-in-go-to-web-page.png" alt-text="Configure the parameters of the Go to Web page action.":::
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/select-parameters-in-go-to-web-page.png" alt-text="Screenshot that shows how to configure the parameters of the Go to web page action.":::
+
+### Resolution 3
+
+Use the Launch Browser action to start the process of the corresponding browser. Then a combination of an additional Launch Browser action with mode set to **Attach to running instance** and a **Wait** action set to a duration afterwards can be used to resolve the issue.
+
+1. Insert a new Launch Browser action as:
+    - **Launch new Microsoft Edge**
+    - **Launch new Chrome**
+    - **Launch new Firefox**
+    - **Launch new Internet Explorer**
+
+2. Set up the parameters of the action:
+    - Launch mode: Launch new Instance
+    - Initial URL: A default URL
+    - Rest parameters can be set as desired.
+
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/parameters-of-launch-new-browser-action.png" alt-text="Screenshot that shows how to set up the parameters of the Launch new browser action.":::
+
+3. Insert an error handling policy:
+    - Select the **On error** option in the action window:
+
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/insert-error-handling-policy-for-action.png" alt-text="Screenshot that shows how to insert an error handling policy by using the On error option.":::
+
+    - Select **Continue flow run** > **Go to next action** in the dropdown list, and then select **Save**.
+
+      :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/save-settings-on-error.png" alt-text="Screenshot that shows how to select the Continue flow run and Go to next action options and then save the settings.":::
+
+4. Insert a **Wait** action and set its duration to **90** seconds (adjust duration as needed).
+
+     :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/wait-action.png" alt-text="Screenshot that shows how to insert a Wait action with value set to 90.":::
+
+5. Insert another Launch Browser action for the same browser used in step 1, and configure the parameters:
+    - Launch mode: Attach to running instance
+    - Attach to browser tab: By URL
+    - Tab URL: The URL inserted in step 1.
+    - Variables produced: Replace the new browser variable with the name of the variable produced in step 1.
+
+        :::image type="content" source="media/failed-to-assume-control-of-edge-chrome-firefox-error/replace-variables-produced-browser.png" alt-text="Screenshot that shows how to replace the new browser variable with the name of the variable produced in step 1.":::
+
+## Cause 3: The browser is launched using a different system user than the one used to run Power Automate for desktop (version 2.38 or higher)
+
+This issue can occur in attended and unattended desktop flow modes.
+
+To resolve the issue, ensure that the browser is launched using the same system user account that's being used to run Power Automate for desktop.
+
+## General checks
 
-Finally, find some general checks that you may apply in case the above actions don't fix the issue:
+If the preceding actions don't resolve the issue, ensure the following:
 
 - Make sure that the environmental variable `ComSpec` with value `C:\WINDOWS\system32\cmd.exe` exists on the machine.
-- Only for Chrome: Check and set the `exit_type` parameter to `normal` at _%localappdata%\Google\Chrome\User Data\Default\Preferences_.
-- Clear the cache and cookies from the browser manually and restart the browser.
+- (Chrome only) Check and set the `exit_type` parameter to `normal` at _%localappdata%\Google\Chrome\User Data\Default\Preferences_.
+- Clear the browser cache and cookies manually, and then restart the browser.

support/power-platform/power-automate/desktop-flows/browser-automation/file-uploader-click-issues.md

@@ -0,0 +1,38 @@
+---
+title: Press Button or Click Link on Web Page Doesn't Work with File Uploader
+description: Solves issues related to interactive file uploaders in web automation.
+ms.custom: sap:Desktop flows\UI or browser automation
+ms.reviewer: amitrou
+ms.author: amitrou
+author: andreas-mitrou
+ms.date: 04/09/2025
+---
+# Press button or click link on web page doesn't work with a file uploader
+
+This article provides workarounds for scenarios where interacting with a file uploader on a web page using automation actions fails.
+
+## Symptoms
+
+When you use the [Press button on web page](/power-automate/desktop-flows/actions-reference/webautomation#pressbuttonbase) or [Click link on web page](/power-automate/desktop-flows/actions-reference/webautomation#clickbase) action in web automation, the action doesn't work as expected if the button or link on the web page is part of a file uploader. The following HTML element is an example of a file uploader:
+
+```html
+<SPAN style="font-size: 14px;font-style: normal;font-weight: 400"><input type="file" /></SPAN>
+```
+
+## Cause
+
+[Web automation actions](/power-automate/desktop-flows/actions-reference/webautomation) create or modify events through scripts, which inform the browser whether the event is trusted. For more information, see [Event: isTrusted property](https://developer.mozilla.org/en-US/docs/Web/API/Event/isTrusted).
+
+For security reasons, browsers might restrict the simulation of events that trigger file upload operations. This restriction ensures that file upload actions are initiated by the user directly rather than through programmatic JavaScript events.
+
+As a result, the **Press button on web page** and **Click link on web page** actions might not work when interacting with file uploaders.
+
+## Workaround
+
+Try one of the following workarounds:
+
+- Use the [Press button in window](/power-automate/desktop-flows/actions-reference/uiautomation#pressbutton) action to simulate user input, bypassing the browser's event validation.
+
+    Instead of using a web element captured from a web automation action or a web browser recorder, capture the element using the **Press button in window** action or through a desktop recorder.
+
+- Use the **Click link on web page** action with the [Send physical click](/power-automate/desktop-flows/how-to/send-physical-clicks-web-element) option enabled.

support/power-platform/power-automate/desktop-flows/browser-automation/js-not-executed.md

@@ -0,0 +1,36 @@
+---
+title: Run JavaScript Function on Web Page Action Fails to Execute
+description: Solves an issue where the Run JavaScript function on web page action isn't executed and returns no error.
+ms.custom: sap:Desktop flows\UI or browser automation
+ms.reviewer: amitrou
+ms.author: amitrou
+author: andreas-mitrou
+ms.date: 04/09/2025
+---
+# The "Run JavaScript function on web page" action doesn't execute but no error is reported
+
+This article provides a resolution for an issue where the [Run JavaScript function on web page](/power-automate/desktop-flows/how-to/populate-text-fields-click-links-javascript) action fails to execute the JavaScript code and doesn't indicate any error.
+
+## Symptoms
+
+The JavaScript code provided in the **Run JavaScript function on web page** action isn't executed, but no error or exception is reported.
+
+## Cause
+
+This issue can occur when a third-party web extension installed and enabled in the browser blocks the JavaScript code execution initiated by Power Automate for desktop.
+
+## Resolution
+
+To solve this issue:
+
+1. Temporarily disable all third-party web extensions in the browser, except the Power Automate extension, to determine if the issue persists.
+
+   - [Turn off an extension in Microsoft Edge](https://support.microsoft.com/en-us/microsoft-edge/add-turn-off-or-remove-extensions-in-microsoft-edge-9c0ec68c-2fbc-2f2c-9ff0-bdc76f46b026)
+   - [Turn off an extension in Google Chrome](https://support.google.com/chrome_webstore/answer/2664769)
+   - [Disabling extensions in Mozilla Firefox](https://support.mozilla.org/en-US/kb/disable-or-remove-add-ons#w_disabling-extensions)
+
+1. If disabling third-party web extensions isn't feasible, you can replace the JavaScript code with a specific web automation action that involves physical interaction. For example:
+
+   If the JavaScript code is designed to click an element using the `document.getElementById('elementID').click();` code, this code can be replaced by the **Click link on web page** action with the **Send physical click** option enabled.
+
+[!INCLUDE [Third-party disclaimer](../../../../includes/third-party-disclaimer.md)]