Recently active topics

If your Agent Studio plugin used to respond in 2 seconds and now takes 8, you might not have a bug, you might have an architecture problem.When you stack multiple actions back-to-back in a conversation process, the AI has to stop and process everything after each one. More actions, more data, more thinking, and your user is stuck watching a spinner.The fix: either add a natural user interaction between actions (like collecting a required field), or bundle your chained actions into a single compound action that runs behind the scenes.One reasoner turn. Cleaner context. Faster plugin.📬Want more tips like this? Subscribe to the Agent Studio Developer Newsletter in Community or via email.

 Hi , I am sending the teams nudge using moveworks send_message API, the nudge only works for 1 email id not for multiple. can someone please help here. following is the code for the reference.        var recipientEmail = event.parm1;// value being passed in event.parm1 is -> email1@example.com,email2@example.com        var summary = event.parm2;// nudge works with following hard coded values    //    emailRecipient = ["email1@example.com", "email2@example.com"];// following code is to get the list of string        var colSplit = recipientEmail.split(",");// logging result is [colSplit 1 -- object] gs.log("colSplit 1" + "--" + typeof colSplit); var arr = []; for (i = 0; i < colSplit.length; i++) { arr.push(colSplit[i]); }       try {            var moveworksAPI = new MoveworksApiSdk();// nudge only works for 1 email id, not for multiple            moveworksAPI.send_message(arr, summary.toString(), event_id);        } catch (ex) {   

Hello Moveworks Community! We are thrilled to announce a major expansion to the Moveworks Data API. Starting April week 1, 2026, we are evolving the API from a "user-engaged" model to a "system-wide" model by introducing visibility into System-Triggered Events.Currently, our Data API focuses on conversations where a user has actively participated. But we know that much of the magic of Moveworks happens proactivity—through notifications, ticket interceptions, and channel resolutions. This update brings that proactive power directly into your hands for deep analysis and reporting. What’s Changing?We are updating four core entities to include System-Triggered Events (Notifications, Channel/Ticket Interception reach-outs)Please note that these changes are ADDITIVE in nature and will not break any of your dashboards. However, to make maximum out of this new data you must update your data pipeline. Conversations Entity: The scope is moving from "user-engaged" to all conversations, including

Hi Team,We are developing a use case to change the profile picture. This is a pretty straightforward use case where we have a File data type slot and an API call to upload the photo to Workday.Workday does have a couple of validations for an uploaded photo that we are trying to check via Moveworks:1. Check whether the uploaded picture is an image (jpeg, png) - we are doing this by checking the file name in the validation of the slot.2. Check the file size. (probably less than 1MB)3. Check if the image is a square or a certain aspect ratio.4. Can we validate the optical recognition - if the uploaded picture is a passport, a driver's license, etc.?I asked the Agent Architect bot, but the recommendation it gave wasn't possible (a script action for the file size).

I have a notify step that uses RENDER() with a generated input string training_str. The template uses HTML tags (e.g. <b>) and a training_str built from a $MAP / $CONCAT expression, but the newline characters render literally (you see \n) when the message is sent to Teams. I’ve also tried "<br>" and "<br >" and they also end up displayed literally. How can I generate a string from my inputs that will render new lines appropriately in Teams? MS Teams screenshot of how the message renders with a literal \n Notify YAML (relevant parts only): - notify: output_key: notification recipient_id: data.user_mw_record.user.record_id message: RENDER(): args: training_str: response_element.t.$MAP(t => $CONCAT(["*📖 ", t.nm, " -- Due ", t.dt, "*"])).$CONCAT("\n") template: >- <b>📣 Notice: Upcoming My Learning Trainings</b> Hi {{recipient_name}}! This is an automated notice that you have <b>{{tr

Hey builders,Steve dropped the Structured Data Analysis GA announcement, then the session turned into a resolver strategy clinic with invoice lookups, custom object slots, and an LLM-powered ticket routing architecture. Ryan co-hosted, Steve did the heavy lifting.What we covered:Structured Data Analysis (SDA) Now GA — code interpreter for large datasets, auto-triggers at 7K tokens Resolver Input Arguments — fixing dynamic resolvers that return the entire table Custom Object Slots — collecting structured data like attendee email + required/optional flag Output Mappers — stripping raw API noise so the reasoner shows only what users need LLM Actions for Ticket Routing — full webhook-to-assignment architecture with generate structured value New Agentic AI Docs — LLM fundamentals, context windows, and reasoning engine pages shippedHere's the recording: Recording linkResourcesStructured Data Analysis Resolver Strategies Context Passing into Dynamic Resolvers Data Retrieval Cookbook — Paginat

I have a workflow where a Compound Action presents a button to the user. Clicking this button triggers a Conversational Process via a Data Mapper payload. Inside this process, the bot asks the user for an input (e.g., a description or amount_to_add slot) and eventually reaches an HTTP action with a Consent card.The Issue:If the user decides to cancel at the consent step (by clicking a Cancel button or typing “cancel”), the flow ends. However, if they scroll up and click the original button to restart the process, the bot remembers their previous answer due to session caching and skips asking the question again.What I’ve Tried:To force the bot to start fresh, I tried passing an empty value for the slot directly from the trigger’s Data Mapper, but ran into blockers:Passing null unquoted resulted in an “Invalid Data Mapper” YAML schema error. Passing “null” or an empty string “‘’” allowed the Data Mapper to save, but clicking the button resulted in a generic error in Slack with zero execu

I noticed a thing, why some plugins have the option to ‘Install’ while in others I have the option to ‘Build’ ‘Get Notified’?     

Hello Moveworks community, We're excited to announce a new troubleshooting tool in Moveworks Setup: API Audit Logs. Partners and customer admins now have full visibility into every external API call Moveworks makes on your behalf — including the URL, HTTP method, response code, and full request/response payloads. (Only for built-in plugins) What Are API Audit Logs?Every time Moveworks calls an external system — your ticketing platform, identity provider, software provisioning tool, or live agent system — that call is now logged and surfaced in real time. Logs are captured near-instantly, as soon as a conversation or workflow execution completes.You can find them at: Moveworks Setup → Organizational Details → Audit → API LogsNote: API Audit Logs cover external API calls made by Moveworks built-in skills only. For Agent Studio workflows, you can refer to Agent Studio Logs. What's CoveredLogs are captured across eight skills:Ticketing — filing, resolving, reopening, and commenting on tick

 Hello Moveworks community,  To prepare for upcoming enhancements to our Moveworks Setup tool —including highly requested features like Configuration Delete—we will be performing a scheduled system migration on Friday, March 6th.To ensure data integrity during this update, we will implement a temporary edit lock on specific configurations. 🗓️ Migration Window The edit lock will be active on Friday, March 6th, during the following times:Region Time Window IST (India) 9:00 AM – 6:00 PM PST (Pacific) 7:30 PM (Mar 5) – 4:30 AM (Mar 6) EST (Eastern) 10:30 PM (Mar 5) – 7:30 AM (Mar 6) CET (Europe) 4:30 AM – 1:30 PM AEST (Australia) 1:30 PM – 10:30 PM  🛠️ Impacted ConfigurationsDuring this window, users will be unable to edit or save changes  for the following: Rich Ticket Forms (RTF) Ticket Workflows Handoff Configurations Routing Configurations Note: This maintenance only impacts the ability to edit these configurations. All existing forms, workflows, and routing l

Hi,I received the attached “Page Not Found” error indicating I need access to the latest version of Agent Studio to complete my Moveworks Agent Studio Developer: Level 1 Certification.Could you please guide me with the next steps to get access.Thanks,Charlene

We recently hosted a session on how to move from AI ambition to real enterprise automation — and what it actually takes to deploy AI agents that work across the business.Here are the key takeaways 👇 🔎 The Core ChallengeEnterprise work rarely lives in one system.Even simple tasks often span:• Multiple applications• Several approvals• Manual handoffs• Status chasing and follow-upsThe result? Operational friction, slower execution, and teams stuck in low-value work. 🤖 What Makes AI Agents Actually ValuableAI agents deliver real impact when they can:• Understand intent• Plan multi-step workflows• Take action across systems• Respect governance and user permissions• Provide a single conversational front doorIn the session, we broke down how the Moveworks reasoning engine works under the hood:It determines who the user is → builds a step-by-step plan → selects the right plugins (actions) → executes → validates the result.Instead of hard-coded workflows, agents dynamically compose “atomic”

Hi all, We have our prod environment set up with Moveworks configured to have a Live Agent Handoff with Glia. We want to make sure our Sandbox is configured properly too. We have a site ID that we inputted as well as a queue ID, but we’re unsure of what our Webhook Token should be - if it’s generated already, how to generate it if it isn’t, or how to retrieve it if it already exists. Do we need to setup/configure a connector? Right now when a live agent handoff attempt is made in our Sandbox environment, the live agent interaction/request is not seen on Glia’s agent side so it seems that it is not reaching Glia at all, and the response on our Moveworks bot is “I attempted again to start a live agent chat for you using the description you provided; the request first went into a pending state and then failed with the internal error LIVE_AGENT_UNABLE_TO_START_CONVERSATION, meaning the system was unable to open a conversation with a live agent.”Does anyone have thoughts on how we can resol

In Analytics, we review the Unsuccessful Knowledge plugin to locate gaps. We noticed that when a form is served, we received an “unsuccessful” for knowledge. Is anyone else seeing this? What are you doing to remedy this issue? 

I’ve setup a Rapid Global notification when a user signs in or out with Webhooks.The challenge I have is I don’t want to enable the notification for everyone, however, Webhooks doesn’t allow you to select an audience to restrict, the Audience option is disabled.It would be useful to be able to control the notification audience on the Moveworks end rather than have to adjust 100s of users in Rapid Global.  

Has anyone integrated with Navan (https://navan.com/)? We are hoping to be able to do travel booking and other tasks in Navan through our AI assistant.