Developers Forum

Hi all,  I am looking to build a child bot that I can insert into a Document Automation Process. I want the child BOT to make decisions based the selection made during validation, ie:wrong document type cant read document other: <free text>I have spent a LOT of time today trying to pass the ValidatorUI output to the child BOT but in all cases the dictionary that is passed down has no records. There must be way to access to this data, any feedback would be much appreciated. For bonus points: Is there a way to create additional rejection options? 

I am not able to capture object with changeable I Frame, every run the I Frame changes which make the capture fails. Note: The portal I use made by MS Dynamics

Hello I'm having trouble deploying bots. If I set it to Regular, it works, but when I try to set it to RDP, which is what I want, I get this message: "User logged into console. Please log out before trying another multi-user deployment."I have the cloud version of Control Room, so I have the latest version.If I uninstall the agent and delete all the folders, then reinstall the agent, it works fine, but when the device logs off, the deployment fails again.

Hey, Good Morning! Couple of days ago, I asked about live logging feature on AA. I kind of found a workaround on how to view live logs (especially business). Here’s my approach, let me know if there’s a way to improve it. We trigger 3 bots simultaneously, in the pool of 6 machines via Jenkins. Each bot needs to create a logs file for the current run, ex - Bot_1_machine_1_timestamp.txt This particular file will be created in a shared folder, accessible to all the bots, in all 6 machines. Once this is created, an email will be sent to business with the below content, where they can click the PowerShell file, which opens the terminal with live logging.!--scriptorstartfragment-->@echo offpowershell -command "Get-Content '\\SharedDrive\Logs\Bot_1_machine_1_timestamp.txt' -Wait"pause!--scriptorendfragm

Hi everyone,I have a bot runner hosted on AWS Workspaces. We decide to switch to RDP from Regular deployment setting because on Regular the screen resolution would be reset to an inadequate resolution.However, on random occasions this error occurs : Unable to perform rdp for the user as we have exceeded the maximum session count. This may be due to multiple reasons. To continue, please Make sure you logout from sessions that are not required at the moment. Then, try to run the bot again. If you continue to see this message, please contact your system administrator.To solve this I have tried to Log off the bots after a run (see screenshot)This error does not occur again but a new one now appears : Bot Launcher for user session **** crashed.Last recorded command ****The last recorded comman

Existing model connections not displaying rather than User Authentication token failed but its worked earlier.   Does any one facing the same issue ? ​@Matt.Stewart ​@Oli.Morris Could you please suggest → if recycle the AA services on all nodes will solve this issue?

Hi,I have reviewed the following:https://ai-kb.automationanywhere.com/agents/external-integrations/confluencehttps://ai-kb.automationanywhere.com/knowledge-base/kb-connectors-confluenceAnd unable to connect to our cloud hosted Confluence instance at all. Using the connector, I received this message:This app has requested Confluence API & Confluence API & User identity API scopes that have not been added to the app. Configure the app at https://developer.atlassian.com/apps and add the following scopes to the Confluence

We are currently working on a PDF extraction use case in Automation Anywhere and wanted to discuss the best approach for implementation.Our current observation is that the native “PDF Extract Field” action appears to rely on coordinate-based extraction. While our PDFs follow a fixed format, there can be minor spacing variations across documents, and we would like to understand the robustness of this approach and the recommended best practices for handling such scenarios.We would appreciate connecting with a Solution Engineer to discuss:Reliability of the PDF Extract Field actionHandling spacing/layout variationsBest practices for fixed-template PDFsPossibility of integrating custom script-based extraction where applicable

I have a customer that is presenting users with a simple form, with two Select File boxes. The user selects the two files, and the next step in the process is a Task Bot that takes actions on those files. Both files are on accessible network shares. The two file objects are being passed to the Task Bot, but attempting to manipulate them ends in very vague failure.Trying to simply cast the file object with .file:getName or .file:getPath returns an error (File name not found, or File Path not found)Attempting to use Process Composer > Get Storage File also fails, but the error message shows the path to the file in CoPilot Storage: <Control Room URL>/storage/v1/file/<big long string for file>I also tried with the Get Storage File v2, action. Unlike v1, it expects a String path to the file and outputs back to a file object; I tried to cast the file object same as above, but no dice.I have been reading through t

In my research it doesn’t seem like it is possible to filter emails by attachment extension or name before downloading them using the out of the box email/outlook packages. I wanted to ask here if there was perhaps another way that I am overlooking.Its possible using graph API so that is my back up if all else fails.

Hi,I’m relatively new to using the AI Skill package. I was solving some practices from the Agentic Quest, and using the package AI Skill: Execute package i got the next error:Last action:GenAIPromptCommandError message:There was a problem at line 1.This may be due to the following reason:Prompt File Path defined in Automation is not found.Location: RFQ (line 1)To continue, edit the bot and fix the error. Then, try again.If you continue to see this message, please contact your system administrator.Code: ba.bot.execution.errorWhat I tested so far:I created the Model Connection using my own OpenAI credentials. I also tested using the Automation Anywhere provider with GPT-4o. The `Test Connection` feature works correctly. Using `Get Response` dir

Hi Team,I am currently working on a use case that interacts with multiple web portals. I have noticed that every 2–3 weeks, a few recorder actions fail due to random DOM XPath changes.Is there an effective or recommended approach to handle such frequent UI changes in Automation Anywhere? Any best practices or suggestions would be greatly appreciated.Thank you!

Just wondering if there is the possibility of connecting A360 Bot Insights to Qlik Cloud? I see that there is a Power BI connector, but no connector to Qlik.

Hi Team,I am using AA360 use case, Where I want to update the records if there are consecutive exceptions then status to ON_HOLD.As UpdateStatus endpoint is not valid anymore which API should be used?Please suggest.Thank you,Prathamesh Dhavale.

Hi Team,I am working on an Automation Anywhere use case that interacts with multiple web applications and portals. One of the major challenges I am facing is handling random dynamic pop-ups that appear unpredictably during execution.These pop-ups can include notifications, session timeout warnings, announcements, update messages, cookie consent banners, or validation alerts. Since they do not appear consistently, the bot sometimes fails or gets stuck when trying to interact with the underlying UI elements.I would like to understand the best practices for handling such dynamic pop-ups in Automation Anywhere.Specifically: How can we efficiently detect unexpected pop-ups during runtime? Is there a recommended design approach to make automation resilient against such UI interruptions? Should we use conditional checks, error handling, object cloning strategies, or separate monitoring tasks? A

I am trying to automate a scenario where I need to perform a drag-and-drop action (for example, dragging an appointment to the service calendar).However, I could not find any direct drag-and-drop command in Automation Anywhere.Could anyone please guide me on:Is there any built-in command available for drag-and-drop?If not, what is the recommended approach to achieve this?Are there any best practices to make it more reliable (avoiding coordinate-based failures)?Any examples or suggestions would be really helpful.

Hi, anyone tried to use the WLM API to add worktimes to the queue using the fil?https://docs.automationanywhere.com/bundle/enterprise-v2019/page/wlm-api-supported.html#tag/Workitems/paths/~1queues~1%7Bid%7D~1file/postThis API needs the binary encoded format of the file to upload. Can anyone post the request body sample to be used in BOT to achieve this? 

Hi, has anyone been able to upload work items from a csv file to a queue via the WLM (Workload Manager) API? If so, could you provide the request body (for when using the REST Web Services : Post Method action). I can perform this task from the Swagger UI, but not through AA actions. Thanks,Laurent

I am using AI to extract text from a handwritten form, I have build the workflow: AI call to extract data Generate a new pdf with the data extracted from AI Merge the original (handwritten) PDF with the newly generated pdf pass the merged document into a learning instance to allow for field / document rules to validate the AI dataMy problem is that the learning instance keeps wanting to get data from the 2nd page. Is there a way force either the learning instance or extraction task to only process the first page of a file? 

I’m using the latest Package SDK (A360-package-sdk-2.13.0.zip) in IntelliJ, Java 11, Gradle 7.6.2.I’ve created packages in the past using an older SDK, they all import just fine.I was wanting to start using the latest Package SDK. So today when I start from the latest Package SDK. It compiles perfectly. Then I get a package.invalid.signature error when importing the new package into the Control Room?Any ideas?

Hi everyone,When executing SQL queries through the Database action in Automation Anywhere, I sometimes only receive a very generic error message such as "syntax error" without any detailed database exception or Oracle error code.Is there any way to view the actual DB error message or stack trace from Automation Anywhere?For example:More detailed Oracle exceptions (ORA-xxxxx) JDBC driver errors Database execution logs Bot Agent logs related to Database actionsI would like to troubleshoot SQL issues more effectively, but the current error messages are too generic to identify the real cause.Are there any recommended settings, logs, or debugging methods for this?Thanks in advance.

Hi Team,I am trying to perform actions using Interrupt Handler package but I am not able to use it as in the handle drop down nothing is coming or showing up in the drop down.Any support or guidance would be of appreciable help. Regards,Abhishek

Using Automation 360, I am trying to click a hyperlink. I can find the hyperlink when using the recorder action and it picks up all the relevant attributes (A tag, innerHtml etc). But when the bot runs, it cannot find the link - throws exception.The hyperlink is in a table that is presented in an iframe.The ultimate aim is for the hyperlink to be clicked within a loop where the bot goes row by row clicking the hyperlink on each row and performing some actions.However, even in debugging to just get the single hyperlink, the bot fails to find the link.Any help would be appreciated..

I have provided all the connection details for Gmail. But it’s giving me below errorError : There was an error in sending the email message. Please check your connection details provided and try again. Please help

En la ultima actualizacion de AA ya no se puede visualizar el numero de linea en el que estas desarrollando cuando pasas las 99 lineas.