Agent Studio

We’re bringing together developers, innovators, and AI enthusiasts for a full-day hackathon at the iconic Chase Center in San Francisco on Tuesday, October 7, 2025. This one-of-a-kind opportunity to build right on the court puts you at the center of innovation where champions compete.✨ Special for Moveworks customers: Bring 2 or more developers from your org, and you’ll be entered for a chance to win a pair of the cutting-edge Meta AI Sunglasses*: smart glasses that bring AI right into your everyday life.Whether you’re building your first AI agent or pushi

I have a list of questions and answers in a restricted google document. It is in a sheet, but it can be moved to a doc as sheets do not seem to be supported. The use case is to ask the bot many different questions that are included in the document and have the bot return with the exact answer for each one of them. Currently this mostly works, but the answers are getting summarized. The answers need to be 100% exact as they are approved internally as correct. If I tell the bot to return the exact answer, it works most of the time, but is there a way to have it always return the exact answer instead of having to ask every time?

Is there a workaround for the output mapping result limit? I'm working on a flow where a user can query the bot to return all of their tickets in JIRA, but I'm losing some data in the process when the mapping gets bigger. Using conversational plugin, not compound action.Example Output Mappings:Mapping 1MAP(): converter: key: item.key project_name: item.fields.project.name status: item.fields.status.statusCategory.name summary: item.fields.summary url: $CONCAT(["https://company.atlassian.net/browse/", item.key]) items: response.issuesreturns:1. Project A – 38 tickets2. Project B – 1 ticket3. Project C – 1 ticket4. Project D – 13 tickets5. Project E – 32 ticketsMapping 2MAP(): converter: assignee: item.fields.assignee.displayN

When working with the Agent Studio 2.0 and I find that I am unable to edit slots after they are created when building decision policies. I am not sure if I am missing something or this is just a limitation of the new Agent Studio. I find that once the slot has been created, I cannot go back and adjust the data type, slot retrieval details, etc. This would be a hinderance when building plugins as a new slot would have to be created any time a necessary change is needed.   

Share in the comments!

Hi All,I am looking at removing a button presented in a notification to the user.Specifically I want to remove a button so the user only has two options. The button I want to remove from the notification shown is “Edit request” to avoid user confusion.Currently selecting “Yes” instructs the middleware to execute a remote action against the machine and clicking “Yes” again will fail as the remote action cannot be executed more than once on the middleware system.I cannot see anything in the CREST Event workspace that would refer to what looks like a consent option to proceed with the choice as shown above.