Developers Forum

I have using recorder to set text in textbot and value is credential type when setting to textbox the value is incorrect.Example: MyUserName value that set to text box “mYuSERnAME”  

Agentic processing and cloud side processing does not allow access to local resources. How can one do detailed process logging when the bots (tools) doing the processing cannot log to a local file or database.My existing bots do a lot of logging, writing to text files and databases. What do you fine people do for debug logging?I do not want to lose the detail that currently have. But I would also like to unleash my bots as cloud running tools so that scripted processes can use those tools or an Agentic process can dynamically pick and choose the tools to do its processing. 

Hi everyone, I have created a bot that runs an automation on Google Chrome. In the beginning, I run the command “Browser:Open”, and then proceed with the recorder.It has run successfully at least several hundred times, and today I got the error “Could not find last opened browser window within the supplied timeout value” → in the try/catch the ErrorLineNumber correspond to the Brower:Open action. In the screenshot taken in case of error the VM desktop was “empty”, without any open window.Before every run, there is a Window : close all command, to close all previously opened windows on the VM. I was unable to reproduce the error, does anyone know where it could come from ?

Hi Team. Why is there no “Enter” in the keyboard? 

I am getting a supplier’s invoice document consistently extracted incorrectly. The invoice PO number is listed as 20493307.The extracted text is always 00439307.The 2 becomes a 0 and the 93 becomes 39.The rest of the text in the description field is correct. only the PO has this issue. When I view this in validator, the validation recognizes the correct value for that field.Any tricks I can try for fixing this?I have tried swapping the OCR engine in my learning instance, but it doesn’t appear to have any affect. 

 <a href="javascript:alert('XSS Found')">XSS</a> 

Good Morning! I’m going through the AA university training videos and trying to learn document processing. I created a learning instance and uploaded the sample invoices provided. It’s been 10 mins, the 1st invoice still shows in progress.My question is - Does it actually take this much time while processing in production or it’s taking more time since it’s the first time I’m triggering the learning instance?Should I keep my device on until it’s completed? I observed in the training, there was a small pop-up as “Bot Running” from the device, so I believe it needs the device to process?  

Hello,I have a pdf with a table containing employee details and the table extends in multiple pages having header in all the pages. Document automation is extracting incomplete data. I have created the learning instance using User defined document type and also tried to use the table prompt as “find table with title 'The List Of Employees' extended in multipage. Extract table details from all pages.” Please let me know if anybody has encountered similar issue and how do we get the entire data accurately.

Previously, the end‑to‑end process for provisioning applications and permissions was fully manual. Users would submit access requests through ServiceNow, which then required approval within Okta. After approval, the corresponding application tile became visible on the user’s Okta dashboard. However, an administrator still had to manually create the user within Automation Anywhere (AA) before the user could actually access the platform.This workflow involved multiple handoffs across different systems and was expected to be managed through Okta Identity Governance (OIG) using SCIM for automated lifecycle management. Since Automation Anywhere does not support SCIM, the question is whether alternative methods exist to achieve a similar automated provisioning flow.

I cannot figure out what is going wrong here.I have a python script which is supposed to return a string value to my task.The basic function returns like this, which combines numbers into a pipe delimited string using f-strings.return f"{total}|{success}|{validation}|{error}"When I run this bot in debug, it works fine.When I deploy this, the function doesn’t fail, but the returned string is “null” instead.“Python output for get_status_counts: null” If I modify my script to write this string to a file, it writes what I would expect. Multiple different bots are having this error. I tried updating packages but it doesn’t seem to have any impact. Part of the log from when this runs:2026-Mar-03 Tue 11:50:14.756 DEBUG [BotRunner-570644a3-2854-43c5-a058-7a6d59646bb9_3e4c19e21d7454cd] - com.automationanywhere.python.commands.ExecuteFunction {deploymentI

I have a datatable as 1*10, but i make the first row as header, then it becomes 0*10. I did this just to specify the column names of dataTable.Also i need to add empty row in this datatable. So I created a record variable that has same column namesThis is 0*10, 0 row, 10 columns, 0.indexed row will be header actually, i just made a screenshot of previous action before making it header.tAttachmentComment variable:this is the record variable which i make empty values:rNoValues variableWhen i try to insert row in datatable as here: 

Hi everyone, I have created a master bot that gets several customer commands to process.This master bot adds to a work queue one element for every command. For example if there are 3 commands, 3 elements will be added to the work queue.I have set the priority of the master bot to low and my expectation was that one : every unit process is processed and only afterwards the master bot is run.However, I have noticed that the master bot takes priority over unit processes. As it is still pending on the customer side, I have a problem of double bot run.Does anyone know how to avoid this problem and always have the work queue bots running BEFORE the scheduled bots ?I hope I am being clear, do not hesitate to tell me if that needs precision.

yesterday i was able to see file uploading system, but there is no now.

Hello, we have lost access to Automation Anywhere since approximately 5 PM.We use the Enterprise version and manage banking services.Could you please tell us if this is a malfunction of AA360 or an internal problem?Thank you. 

I use a server for the Control Room in a corporate environment, and we are receiving recurring alerts of excessive RAM usage, caused by multiple instances of Zulu Platform x64 Architecture, used by Automation Anywhere's CR.How could you help with this? What could be causing this excessive increase, and how can we resolve this situation? The server in question already has 48GB of RAM.

Hi, I am using outlook and I want to connect to a shared mailbox with A360, so I can read emails from that mailbox. I see in some posts on this forum that that functionality isn't available (yet?). Does anyone have a workaround for this problem?

Hi All ,I’m trying to use function keys using terminal emulator command Send key but behavior of key functionality is different in terminal emulator.For example if bot is sending the key - PF2,PF8 its typing OQ & [19 .connection type : VTSERIESplease help me on this.

I need to automate hierarchy structure.structure design is.Company      ManagerName            User name(Cut Option)      Deactivated Manager Name(Pasteunder the username which i cut)

Hi All,We are looking to automate the actions of a tool that requires us to perform a mouse drag-and-drop operation. The tool requires us to click the mouse at a specific coordinate, hold it down, and drag it to another coordinate. We cannot use keyboard shortcuts, only mouse clicks based on coordinates. Is this possible to achieve using AA?

Hi All,We are currently facing an issue with the Email Event Trigger in Automation 360.When a new email arrives, the Email Event Trigger fires correctly, but it is automatically marking the email as “Read”. Because of this, when the bot starts and looks for unread emails, it does not find anything — the trigger has already read the email in the first instance.Has anyone faced this behavior before, and is there a way to prevent the Email Trigger from marking emails as read? Or is there a recommended best practice to handle this scenario?Any guidance or workaround would be really helpful.Thanks!

my automation failed with that error, i've make sure the path is correct, the extension correct what could be the problem? i've tried opening the excel file and a pop up appeared it says that something is wrong about the contect, so i added "if window exist" but the problem doesnt go anywhere, it can't even open the excel file how to fix that?

Tried to run the java script using Browser-RunJavaScript action-Its getting timed out.

Hey People!I’m new to AA and was wondering where is version of the packages mentioned in the control room?Quick Update, I received this notification from AA - Chrome is removing the UiAutomationProviderEnabled transition policy after v146 (March 25, 2026). To prevent Recorder web automation failures in Chrome v147, upgrade the Recorder package to v2.17.6+ (includes UI Automation support).When I tried to search the package version of recorder, I could only find the below details. My question is the version mentioned by the AA to upgrade is v2.17.6+, how do I know my current version? I only see 5.5.2-20260205-144056, which means I’m already at v5?

Hi Everyone,In SAP GUI, a report is displayed in table format on the screen. Visually, it looks like a structured table with rows and columns.However, when using Automation Anywhere A360: Object Cloning does not identify it as a table control. It captures only single individual elements instead of the entire grid. When using Recorder, it captures properties of each element separately rather than recognizing it as a full table. Because of this, I am unable to extract the entire table in one go.In such cases, what is the recommended way to extract the complete table data from SAP when the object is not recognized as a proper table control?Thanks in advance.

We use the WLM to run our bots. The bot is linked to the WorkItem Template.  The Queue is also paired with the bot. The bot automatically has the record variable workItem. Is there an easy way to retrieve the queue name and workItem ID when that bot runs? Perhaps a parameter to the variable? I’m thinking this information has to be easily obtainable in some fashion. Any help is greatly appreciated!Jaye