Product Update

Xurrent ITSM - October 2026 Product Updates

Jacob Roscoe
October 29, 2026
32 Min Read

Table of contents

Downward-pointing chevron dropdown arrow icon in black.

This post is a living document, updated throughout the month as new product releases roll out. Each update listed here is released to our QA environment on the date shown and typically promoted to production the following week. Along the way, customers can review changes, share feedback, and help shape what ships next. Xurrent delivers product updates on a weekly cadence to keep enhancements moving continuously.

October 29, 2026

Account Trusts: Limit Approval Assignment to Skill Pools

What this gets you: You can limit which of your people a trusted account can assign approvals to.

The trust option "can assign workflow approval tasks to anyone of us" now has a list of skill pools, "Limit to members of these skill pools (leave empty for anyone)". The list appears while the option is checked. Previously, the option made every person of the granting account available to the trusted account as an approver.

Shows a checked account trust option with a linked list field beneath it that narrows who can be selected.

When skill pools are selected, the "Approvers" field in the trusted account offers only their members from the granting account. Saving an approval with anyone else from the granting account fails, including through the API and imports. People who qualify as approvers in another way, such as the specialists of the task's own account, are not affected. When the list is empty, the option works as before.

Task templates and workflow tasks of the trusted account can use a listed skill pool as their approval skill pool. The trusted account sees the name of the skill pool, but cannot open it.

A change to the list applies to new approvals only. Approvals that already exist stay with their approvers. An out-of-office delegate outside the listed skill pools does not become an approver.

Sera AI: Sera Assist Knows What a Record Is Linked To

What this gets you: You can ask Sera Assist what is related to the record you are on, and get the linked records without looking up a number.

On a request, Sera Assist now knows the request group it is in, the requests grouped into it, and the problem, workflow or project it is linked to. It also knows the records referenced in its notes. On a task, it knows the workflow, the requests of that workflow, and the predecessor and successor tasks. Previously, Sera Assist searched for similar wording and could present loosely similar records as related.

Sera Assist lists each record with its type, number, subject and status, and with how it is related, for example as the linked problem or a predecessor task. Click a record to open it. When nothing is linked, Sera Assist says so.

Shows Sera Assist answering what a request is linked to, naming the linked workflow and the records referenced in its notes.

Sera Assist does not list records as related when they only share a configuration item, a service instance or a requester. To find those, ask Sera Assist for similar records. An answer lists up to 25 records and says when there are more. It lists only records you are allowed to see.

Agent Designer: Instructions for the Intake Agent

What this gets you: You can decide how the Intake Agent writes to requesters, and add details it asks for.

The Intake Agent's page in Agent Designer has an "Agent instructions" block. Use it to set the tone of the agent's message to a requester, the greeting and sign-off, the length, the maximum number of questions per message, and the account's own terms. You can also name details to ask for, on every request or on requests for a given service or category. For example: "For requests on the Laptop service, ask for the asset tag."

Shows the Intake Agent page with instructions for tone, greeting, sign-off and length, confirmed as "Approved and saved."

The block holds up to 1,000 characters. Every saved version goes through a check before it takes effect, and the result appears next to the block: "Approved and saved.", "Partially applied" or "Rejected", with the reason. When instructions are partially applied, the block keeps only the accepted text.

Instructions cannot remove a check the agent already makes, mark a request as complete, or change the language of the message. The agent's internal note on a complete request stays as it is.

An Account Designer can edit the block, and an Auditor can read it. The Agent Designer "Audit Trail" lists each change.

Agent Designer: Agent Runs, Success Rate and Time Saved in Reports

What this gets you: You can see what is behind each Agent Designer tile, look back further than 7 days, and add the figures to a dashboard.

The "Runs This Week", "Avg Success Rate" and "Time Saved/Week" tiles in Agent Designer now open a report. "Agent Runs", "Agent Success Rate" and "Agent Time Saved" show the tile's figure day by day for the same 7 days. Previously, the tiles showed a number with nothing behind it.

From the report, you can widen the period, and filter or group by agent, trigger, outcome or skill. "Show Underlying Data" opens the requests behind a value.

Shows the "Agent Runs" report as a day-by-day line chart for the Intake Agent over the last 7 days.

The three reports also appear under "Analytics" > "Reports", and you can add them to a dashboard. Opened from there, they show the last 6 months by week, with one series per agent.

UI Extensions: Checkbox Lists in Columns

What this gets you: Long checkbox lists take less space on a form.

A multiple-choice field shown as checkboxes now has a "Columns" setting in the UI extension designer: "One column", "Two columns" or "Fit to width". "Fit to width" places as many columns as the width of the field allows. The options fill each column from top to bottom, so they keep their order.

Shows a checkbox field laid out in two columns, with the new "Columns" setting open in the UI extension designer.

The layout applies on the specialist form, in Self Service and in the designer preview. On a phone, the list shows one column. Existing forms keep one column until you change the setting.

Shop: Order on Behalf of the Same People as for Requests

What this gets you: Employees who submit requests for people in other internal organizations can now order Shop articles for them too.

When an account has both "Allow requests to be submitted on behalf of someone else" and "Allow shop articles to be ordered on behalf of someone else" on, an employee of an internal organization can pick the same people at Shop checkout as on a request. Previously, the Shop offered only people of the employee's own organization and the organizations below it.

Nothing changes when only the Shop setting is on. Employees of external organizations still order only within their own organization. The order template and article coverage of the selected person still apply. See Important Messages.

Requests: Follow-Up Reminders Appear in the Notes

What this gets you: You can see in a request's notes when a follow-up reminder went to the customer, without opening the "Audit Trail".

When a request is "Waiting for Customer" and a follow-up reminder is sent, a system note now appears in the notes of the request. It names the person the reminder went to and how many business days the request had been waiting for a response. A rule with two reminder steps adds a note for each reminder. The last reminder adds no note when it also completes the request. Previously, only the "Audit Trail" recorded the reminder.

Shows two system notes on a "Waiting for Customer" request, each recording a follow-up reminder and the business days waited.

Requesters do not see the note in Self Service. The "Audit Trail" still records the reminder.

CMDB: Asset Management Is Now CMDB

What this gets you: Daily configuration work and CMDB setup each have one place.

The "Asset Management" item in the menu is now "CMDB". It opens on the "Dashboard" tab, which shows the same figures as the former "Reports" tab. The tabs are now "Dashboard", "Configuration items", "Reconciliation queue" and "Lifecycle reviews".

Shows the renamed CMDB area opening on its Dashboard tab, with the "CMDB Settings" menu at the end of the tab row.

CI rules, products, product categories and the other CMDB setup screens are now under "Settings" > "CMDB". A "CMDB Settings" menu at the end of the tab row links to each of them. "Lists" no longer includes product categories. Each moved screen keeps its content and actions.

Single Sign-On: Choose the Directory Account's SSO Configuration

What this gets you: You can make sure the people of a support domain sign in through the right identity provider.

A support domain account can sign people in with an SSO configuration of its directory account, through "Login using SSO configuration of directory account". When the directory account has more than one enabled SSO configuration, the support domain's SSO configuration now shows the "Reference" field. Set it to the reference of the directory account's SSO configuration you want the support domain to use.

Previously, the field stayed hidden when the support domain had only one SSO configuration. The support domain then used the directory account's configuration with the reference "primary", or else the first enabled one.

Nothing changes for an account until you change the reference.

iPaaS: Jira Cloud Connector

Xurrent iPaaS now includes a native Jira Cloud connector for automating Jira issue, comment and user workflows. Previously, Xurrent iPaaS had no native connector for Jira Cloud.

The connector supports:

  • Authentication with an Atlassian account email and API token, including service accounts.
  • Creating, updating, retrieving and searching issues, including search by JQL.
  • Updating issue status and assigning users to issues.
  • Retrieving an issue's changelog and field schema.
  • Creating, updating and retrieving issue comments.
  • Uploading, downloading and listing issue attachments.
  • Retrieving user details, creating users and searching assignable users.
  • Triggering workflows in real time when issues or comments are created or updated, through Jira webhooks.
  • Triggering workflows on new or updated issues through scheduled polling.
  • Jira REST API requests for capabilities the native actions do not cover.

Workflows can start in real time from Jira webhooks or on a schedule through polling. Setup steps and the full list of actions and triggers are in the Jira Cloud Connector documentation.

October 22, 2026

Reports: Tag Usage Across Record Types

What this gets you: You can see which tags are used most across all your work, and open the records behind each count.

Analytics now offers reports on how tags are used. "Active Records by Tag" shows the most-used tags in the account, with one bar per tag. Each bar is split by record type: requests, problems, workflows, tasks, projects and project tasks. Previously, tag counts were only shown on each tag's own page, one tag at a time.

"Completed Records by Tag" counts finished work the same way. It always covers a period, and opens on the last 3 months. Each record type is counted by its own completion date.

Click a segment to open the list of that record type, filtered on the tag. The bar chart shows the 20 most-used tags. Switch to "Table" to see every tag, with a column per record type and a "Total" column. Click a record type in the legend to hide it. A tag filter narrows every record type. A filter that only some record types have, such as "Category", narrows only those record types.

Twelve reports for a single record type are also available, such as "Open Requests by Tag", "Active Problems by Tag" and "Finished Tasks by Tag". They open with the filter "Tag is not None", so untagged records are left out. Remove the filter to include them.

The counts cover only the records the viewer has access to. They can therefore differ from the counts on a tag's own page.

Service Level Management: Open Any Month Since the Report Started

What this gets you: You can open an SLA report for any month back to the start of the SLA, even when that is years ago.

The month picker of the "SLA Reports", "Customer Reports", "Provider Reports" and "FLSA Reports" now offers every month from the start of the report to the current month. Previously, the picker listed only the 18 most recent months, so older months looked as if their data was lost. That data was always kept.

The picker shows one year at a time, as a grid of 12 months. Use the buttons beside the year to move to the previous or next year. Months before the start of the report and after the current month are shown, but cannot be selected. The picker opens on the year of the month you are viewing, with that month highlighted.

October 15, 2026

Time Tracking: Time Entry Privacy for Directory Accounts

What this gets you: Recorded time can stay private to the person who logged it and their manager, across every account in your directory.

A directory account can now limit who sees recorded time. With the new "Time entry privacy" setting, specialists see only their own time entries and those of the people in the organizations they manage. Previously, any specialist could read, filter and export the time entries of every colleague.

The setting is on the "Account Settings" page of the directory account, and it is off by default. When an account administrator checks it, the setting applies to the directory account and all of its support domain accounts. The manager or substitute of an organization sees the time entries of that organization's people, but not those of the organizations below it.

Shows the new "Time entry privacy" checkbox on the directory account's Account Settings page.

The limit applies to the Time Entries console, the time entry reports and their exports, and the time entries listed on a request, task, workflow, project task or project. It also applies to the REST and GraphQL APIs. In Resource Planning, time entries a specialist cannot see are no longer counted as "Other Effort". The total time spent on a record stays visible, and the Timesheets console is unchanged.

Account Administrators, Directory Administrators, Auditors, Directory Auditors and Financial Managers still see all time entries. Holding one of these roles in any account of the directory is enough. Once the setting is on, an integration that reads time entries with a specialist's token receives only that person's entries. Use the token of one of these roles for integrations that need every entry.

Self Service: Knowledge Article Library

What this gets you: Your users can browse every article available to them in one place, without knowing what to search for.

Self Service can now offer a list of the knowledge articles a user can open. Users browse the articles across all services without typing a search. Previously, users found articles only by searching or by opening one service at a time.

Shows "Knowledge Articles" being chosen as a Single Navigation Menu Item in the Self Service Design.

The list appears only after an administrator adds it to the navigation. In the "Self Service Design", add a "Single Navigation Menu Item" and choose "Knowledge Articles". The item is not part of the default design or the full navigation menu, so nothing changes in Self Service until then.

Shows the Self Service knowledge article list, with each article's subject and service and a filter field at the top.

Each row shows the article's subject and service, and the list starts with the most recently updated. The search icon in the header filters the list once you type two characters. The list holds the articles the user can find through Self Service search, and opening one counts as a Self Service view.

GraphQL API: Team Involvement

What this gets you: You can pull team handoffs and holding times into your own reporting tools and data warehouse.

Team involvement can now be read through the GraphQL API. Team involvement tracks which team held a request, from when until when, and for how long. It is the data behind the "Team Holding Time" and "Team Handoff" reports. Previously, it was only available in those reports and in the team involvements console.

The new teamInvolvements connection lists the entries of an account, with filters and sorting. Each entry gives the request, the team, the previous team, the member, the start and end time, the duration in seconds, and whether it was a handoff. An open entry has no end time or duration yet. Integrations can filter on updatedAt to fetch only the entries that changed since their last call.

On a request, the teamInvolvements connection returns its chain from the first team to the latest, for the account in the X-Xurrent-Account header. The data is read-only, and there is no REST endpoint or export. Personal access tokens and OAuth applications need the new "Team involvement" scope.

Team Involvement: Holding Time Ends When a Request Is Grouped

What this gets you: "Team Holding Time" counts the time a team held a request, and stops counting once a request group takes over.

Grouping a request now ends its team's holding time. When a request is grouped into a request group, or turned into a request group, its open team involvement entry ends at that moment. Previously, the entry stayed open, so the team kept accruing time in the "Team Holding Time" report, even after the group was completed.

When an active request is ungrouped, tracking starts again with a new entry. Its "From team" is the team that held the request before it was grouped. Team changes made while a request is grouped are still not recorded.

After the release, a one-time correction ends the entries left open on grouped requests. See Important Messages.

Service Level Management: Correcting Finished Work Keeps Its SLA Results

What this gets you: You can fix the service instance on completed requests without changing past SLA scores.

Changing the service instance of a completed request no longer changes its SLA results. The affected SLA that measured the work keeps its times, its scores and the month it counts in. Previously, the correction created a new affected SLA for the new service instance, with zero durations, scored as met in the month of the edit.

The same applies to a provider that has finished its part while the customer is still working on the request. The provider's resolved affected SLAs stay resolved. Accounts that are still working on the request follow the new service instance, as before.

This applies when the service instance is the only change. A new measurement still starts when the request is reopened or "Requested for" changes. It also starts when the category changes to Incident, RFC or RFI, when a team of another account takes the request over, or when the completion reason is or was "Declined". Earlier corrections keep the affected SLAs they created. See Important Messages.

Sera AI: Approve and Reject Tasks in Chat

What this gets you: Approvers can decide an approval task from the Sera AI chat, and nothing is recorded until they confirm.

Shows Sera AI presenting a change controller approval with Approve and Reject options after an approver asks to approve a task.

When an approver asks Sera to approve or reject a task, Sera first shows the decision on a card, and Xurrent records it only after the approver confirms. Previously, approvers had to open the approval to decide it. Sera asks for a reason before recording a rejection, and for a note on approval when the task requires one. The decision is recorded against the approver, the same way as on the approval form. AI agents cannot decide an approval. When an approval needs fields the chat cannot collect, Sera points the approver to the approval itself.

Shows the new "Task Approvals" option in the Sera AI section of the Account Settings page.

The setting is off by default. To turn it on, an account administrator selects "Task Approvals" in the Sera AI section of the "Account Settings" page. The setting shows only when Sera AI is enabled for the account. Until it is on, Sera tells an approver who asks that deciding approvals in a conversation is not switched on, and points them to the approval.

October 8, 2026

Problem Management: Tasks on Every Problem

Problems now have tasks. A new "Problem Tasks" section on every problem lists the work needed to fix it. Each task has a Subject, Instructions, a Type, a Priority, a Team and Member, a Status, a Desired completion and a linked Risk. Previously, a problem had no place to split its work into steps that could be assigned.

Problem tasks are available in every account, with no setup needed. Account Administrators and Problem Managers add tasks with "Add task" in the section. Specialists can update and close tasks, and Auditors can read them. In a Strong Privacy account, Specialists see only the tasks of their own teams.

Shows the new "Problem Tasks" section on a problem, listing assigned tasks with type, status and target date.

Expanding a task shows its instructions and details. Drag a task to change the order, which saves at once. "Close task" opens a short form where you choose Completed or Canceled and enter a close note, which is required. A closed task can be reopened from its own form by setting an open status again.

Shows an expanded problem task with its instructions, type, priority, team, member, risk and the "Close task" button.

A problem cannot be set to "Solved" while one of its tasks with priority "High - Essential to the fix" is open. This applies to every way of solving a problem, including the API, imports and automation rules. It also counts tasks the person solving cannot see. Open tasks with priority "Medium - Expected" or "Low - Optional" do not block. The problem form shows a warning for them instead. When a workflow or project completes while an essential task is open, the problem returns to "Assigned" with a note instead of being solved.

Shows a problem task opened from the member's Inbox, with its linked problem, instructions, priority and assignment.

Problem tasks appear in their member's Inbox, in global search, and in a new "Problem Task" console in the Records menu. Full editing is available from the Inbox, global search and the Problem Task console. Editing from the problem record is in progress, and problem task reporting will follow. A member receives an email notification when a task is assigned to them. Account Administrators maintain the options for the Type field on the new "Problem Task Types" settings page. Problem tasks can be exported and imported, and are available through the REST and GraphQL APIs. See Important Messages.

Problem Management: Complete Linked Requests With the Resolution

Problems have two new fields, "Resolution" and "Customer resolution". When a problem is solved, "Save and complete requests" completes its open linked requests and sends each requester the customer resolution. Previously, the linked requests of a solved problem had to be completed one by one.

Shows the new "Resolution" and "Customer resolution" fields on the problem form.

"Resolution" records the fix for specialists. "Customer resolution" is the text requesters receive, and it is shown once the problem has linked requests. When it is empty, the Resolution is sent instead. A confirmation lists the requests before anything is sent. A written resolution also counts as the note that solving a problem requires.

Shows a problem being solved with both resolution fields filled in and the new "Save and complete requests" button.

"Save and complete requests" is offered while solving and on a problem that is already solved, as long as it has open linked requests. Each request is completed with the completion reason "Solved" and is assigned to the problem's Manager. The text becomes its completion note, posted in your name, and also appears in a new "Resolution" field on the request.

Some linked requests are left as they are. These are requests that are already completed or waiting for a workflow or project, requests grouped into another linked request, and requests reopened after an earlier resolution. Requests linked to the problem only from another account also stay open. Editing the resolution later does not change requests that are already completed.

The agile board dialog for the done column and the "Complete…" action in the Problems console offer the same two fields and an "Also complete linked requests" checkbox. A new account setting, "Require a resolution when a problem is solved", makes the Resolution mandatory, including through the API and automation rules. It is off by default. See Important Messages.

Notes: Mention a Team

Specialists can now @-mention a team in a request note. Typing @ offers a choice between "Person" and "Team". The "Team" view lists the teams you can mention, with their number of members. Previously, looping in a team meant mentioning each member by name.

Shows the "Enable team mentions" checkbox in the new "Notifications" section of a team.

Team mentions are off until they are turned on for a team. An Account Administrator, or the person set as the team's Manager, checks "Enable team mentions" in the new "Notifications" section of the team. Only teams of accounts where you have the Specialist role are offered, so a service provider's specialists never see a customer's teams. In a Strong Privacy account, only teams that can see the request are offered.

Shows the @-mention picker in a request note with the "Team" view selected and a team listed with its member count.

When the note is saved, every member of the team who can already see the note is notified: "<Author> mentioned your team <Team> in the following request". A team mention never gives anyone access to the request. If no other member can see the note, the note is still saved and you see a message that nobody was notified. Each person can turn these emails off with "Do not notify me when one of my teams is mentioned", from the "Notifications Preferences" in "My Profile".

You can mention a team on the request form, in toolbar actions such as Forward and Complete, and on agile boards. Team mentions are not offered where one note goes to several requests, such as "Edit All…". They are also not offered in Self Service or the mobile app. In emails and in Self Service, a team mention reads "@Team Name".

Inbox: Workflows Waiting to Start or Halted

My Inbox now lists the workflows that need your attention. These are workflows with the status "Progress Halted", and workflows with the status "Registered" that have no Start date. Previously, workflows did not appear in the Inbox.

Shows a Registered workflow without a Start date listed in My Inbox alongside a request.

A Registered workflow appears in the inbox of the person who created it, when that person has the Workflow Requestor role and is not its manager. All other entries go to the workflow's Manager, and move when the Manager changes. An entry leaves the inbox when the workflow starts, resumes progress, completes, is canceled or gets a Start date. Workflows never appear in team inboxes.

Workflow entries do not raise the Inbox badge. The "Record type" filter has a new "Workflow" option. Notifications for a halted workflow are unchanged. See Important Messages.

Request Groups: Own Configuration Items and a Major Incidents View

In a request group with the category "Incident", each grouped request now keeps its own configuration items. Previously, grouping a request replaced its configuration items with those of the group. A specialist who may edit a grouped request can change its configuration items, and the change is kept. Groups of other categories are unchanged.

On an incident group, the configuration item field is labeled "Configuration items held responsible". The "Grouped Requests" section shows each request with its configuration items and "reported by" its requester. Requests created with "I'm Also Affected" on the group's broadcast start with the group's configuration items, which can be corrected afterwards.

Shows a major incident group with "Configuration items held responsible" and each grouped request listed with its own configuration item and requester.

The Requests console has a new "Major Incidents" view. It lists request groups that have a major incident status, and opens on the open ones. "Major incident status" is also a new column you can add to your own request views. Previously, the status was shown only as an icon in the Impact column.

A request of another category can no longer be grouped into a major incident, and a major incident can no longer be grouped into a request group of another category. Once an incident group has grouped requests, its category can no longer be changed. Requests grouped before this release keep the configuration items they have now. See Important Messages.

Configuration Items: RAM in Any Unit

The "RAM amount" field of a server configuration item now takes a number and a unit, from bytes to petabytes. The unit starts at GB, so typing 16 still means 16 GB. Previously, RAM could only be entered in GB.

Shows the "RAM amount" field with its new unit dropdown, from Bytes to PB.

The configuration item shows the value in a readable unit, such as "16 GB". Hovering over it shows the exact number of bytes. Editing reopens the field in the same unit, and saving without changing it keeps the exact value. RAM and drive capacity are now shown with up to two decimals, so 15.625 GB shows as 15.63 GB.

Existing values are converted automatically and stay the same. In configuration item views, the previous field is now labeled "RAM amount (GB)". The export has a new "RAM Bytes" column, and an import accepts either column. For integrations, the API has a new ram_bytes field (ramBytes in GraphQL), and ram_amount is deprecated. See Important Messages.

CI Reconciliation: Create Approval Rules From the Review

While reviewing a change in the Reconciliation Queue, you can now create the automation rules that approve similar changes in future. Previously, these rules had to be written by hand.

When no existing configuration item matches, the review offers "Auto-approve new CIs from source <source>" under "Automated CI creation". When no product matches, the "Product not found" review offers "Auto-create products from source <source>". A confirmation explains what the rules will do, and "Create rules" adds them.

Shows a pending reconciliation review with the "Automated CI creation" offer to auto-approve new CIs from the Lansweeper source.

The rules act on new changes from that source. Changes already on hold stay where they are, and incomplete captures still wait for review. A product is created only when a category and a support team can be found for it. The rules appear under Reconciliation Rules and can be edited, disabled or deleted like any other rule.

Once the rules exist, the review shows the rule that acts on that hold and what it did for this change, with a link to the Automation Trail. The offers are shown to Account Administrators, Account Designers and Configuration Managers.

Shop Article Bundles: One-Time and Recurring Prices Together

A shop article bundle can now combine one-time articles with articles that share one recurring period. An example is a workstation bought once and a notebook leased monthly. Previously, all articles in a bundle had to use the same pricing.

The bundle's pricing shows "Price (sum of parts)" and "Recurring price (sum of parts)" with its period. Requesters see both totals on the bundle and in the cart, and the order line records both.

Shows a bundle combining a purchased workstation and a monthly lease, with both the one-time and recurring sum-of-parts prices.

Recurring articles in one bundle must still share a period, so monthly and yearly articles cannot be combined. The "Bundle Items" picker offers one-time articles and articles of the bundle's period. A "Custom bundle recurring price" must use the same period as the recurring articles.

iPaaS: Microsoft Outlook Connector

Xurrent iPaaS now includes a native Microsoft Outlook connector, built on Microsoft Graph, for automating Outlook and Microsoft 365 email workflows. Previously, Xurrent iPaaS had no native connector for Outlook.

The connector supports:

  • Client Credentials authentication.
  • Reading, listing, searching, sending, moving and archiving emails.
  • User and shared mailboxes.
  • Retrieving and downloading attachments.
  • Email content in HTML or plain-text format.
  • Searching and filtering by subject, sender, folder, body, categories, importance, attachments and date.
  • Triggering workflows when new emails arrive, including emails in shared mailboxes.
  • Near-real-time monitoring through Microsoft Graph change notifications.
  • Microsoft Graph API requests for capabilities the native actions do not cover.

Client Credentials authentication requires an app registration in Microsoft Entra ID, with Microsoft Graph application permissions granted by a tenant administrator. Setup steps, required permissions and the full list of actions and triggers are in the Outlook / Microsoft Graph Connector documentation.

October 1, 2026

UI Extensions: Multiple Choice Questions as Checkboxes

A Select field with "Multiple" on can now show its options as a list of checkboxes instead of a list box. Requesters tick the answers they want. Previously, choosing more than one value in the list box meant holding down a key while clicking, which many requesters overlooked.

The new setting is "Show as checkboxes" in the field's settings in the UI extension designer. It appears once "Multiple" is ticked and is off by default. With it on, the form shows one checkbox per option under the field's label, in the order the options are listed. This applies in both the specialist interface and Self Service.

Shows a multiple-choice Select field rendered as checkboxes in the UI extension designer, with "Show as checkboxes" ticked in Field Settings.

The answer is stored exactly as the list box stores it. Filters, reports, views, automation rules, the API and exports are unchanged. The setting can be turned on or off on a UI extension that is already in use, and answers people have already given stay as they are.

"Required", "Minimum", "Maximum" and option visibility conditions work on the checkboxes as they do on the list box. Option labels appear in the reader's language. "Include empty option", "Placeholder" and "Visible items" are not offered while the setting is on. "Show when empty" is offered, so the question can stay on the record when the requester ticks nothing.

Specialists see the question and the chosen answers on the record. Existing forms are unchanged until the setting is turned on in the UI extension.

Dashboards: PDF Export on A4 Pages

A dashboard can now be exported to PDF as a document of A4 pages, with charts and text at normal print size. Previously, a dashboard always exported as one sheet as large as the dashboard, so the text on a wide dashboard became too small to read.

The "Print Dashboard to PDF" dialog and the dashboard export form both have a new "Page layout" choice: "Single page" or "A4 pages". "Single page" is the default and produces the same output as before. Existing export schedules keep "Single page".

Shows the "Print Dashboard to PDF" dialog with the new "Page layout" choice set to "A4 pages".

With "A4 pages", the dashboard is laid out at the width of the page and continues onto as many pages as it needs, in the orientation you choose. A widget that does not fit on the rest of a page moves whole to the next page. A widget taller than a page is split. The title and filters appear on the first page only, and pages are not numbered. A scheduled export uses the same choice on every run.

An account that has changed the widget markup in its dashboard PDF design may see a widget that fits on one page split across two pages with "A4 pages".

Inbox: Agent Actions for AI Agent Suggestions

When an AI agent skill runs in "Suggest", its suggestion now waits for a decision in a new "Agent Actions" view in the Inbox, next to "My Inbox", "My Team's Inbox" and "Watchlist". Previously, a suggestion arrived only as an internal note on the request and was easy to miss. The internal note is still added, and the "Agent Actions" item points to it for the agent's reasoning.

Each item goes to the request's "Member" and to the team that holds the request. Nobody else sees it. The "Agent Actions" view appears only when you have at least one open item.

Shows the new "Agent Actions" Inbox view listing pending AI agent suggestions with impact, skill, confidence and created time.

Six skills send their suggestions here: "Reopen Signal Detection" and "Verification Note Generation" on the Closure Agent, "Duplicate Detection", "CI Linking" and "Impact Assessment" on the Triage Agent, and "Resolution Reply" on the Resolution Agent. A skill running in "Auto Confirm" applies its change directly and adds nothing to the view.

Each row shows the request, its impact, the skill and the agent's confidence. Clicking a row opens the request. When you act on the suggestion there, the item closes on its own.

Sera AI Studio: Service Coverage Gaps on the Insights Tab

Sera AI Studio has a new "Insights" tab. Its first section, "Service Coverage Gaps", lists requests Sera could not route because a support path is misconfigured. Previously, Sera gave the end user a fallback message and nothing told administrators that the path was broken.

Each row shows the service or request template Sera tried to route to (the "Subject" column), the support domain, the reason, the number of occurrences and when the gap was last seen. The reason is one of "Sera AI not enabled", "No service desk team" or "No active support agreement". A matching button opens the place to fix it: "Enable Sera AI", "Assign team" or "Configure SLA".

Shows the "Service Coverage Gaps" section of the Sera AI Studio Insights tab, with each gap's domain, reason, occurrences and a fix button.

A row clears on its own once the misconfiguration is fixed, and comes back if the gap recurs. You can filter by support domain and by period. A row for a support domain you cannot administer has no fix button, because the change is made in that domain's own account. A row shows "Any" when the gap involved no particular service or template. The tab stores no requester names and none of the text requesters type.

The tab is available to Account Administrators, Account Designers and Auditors.

CI Reconciliation: Decide Once on New Product Fields

CI Reconciliation now checks custom field values against the product before applying them. When a record proposes a field that the product's UI extension does not define, the record waits in a new "Product Enhancement" view of the Reconciliation Queue, with the on hold reason "Product custom fields pending". Previously, such values were saved on the CI without being added to the product's form, so nobody saw them.

This applies to accounts with "Enable CI Reconciliation" turned on in Account Settings, and to saves that provide both a Source and a Source ID. That includes API integrations, asset discovery, and CI form saves with the Reconciliation option ticked. Records that carry no custom field values go straight on to review.

Shows a pending reconciliation record listing its new custom fields, with an example value, label, section and Add or Ignore decision for each.

A Configuration Manager or Account Administrator decides each new field once per product. The view shows the product each record is waiting on. Opening a record lists its pending fields with an example value. Choosing Add puts the field on the product's UI extension with the label and section you choose. If the product has no UI extension yet, the first Add creates one, named after the product. The field is searchable and filterable, and internal, so it does not appear in Self Service. Choosing Ignore drops the value, and the question is not asked again for that product. One decision releases every record waiting on the same product. Automation rules in the new Product Enhancement Rules group can react when a record starts waiting.

Add is not available when the product's UI extension is a legacy HTML extension, or has a draft version in progress. Add the field in the UI extension instead, or activate or discard the draft. The waiting records are then released automatically.

A product can opt out with the new "Ignore new custom fields" option on the product. New field values for that product are then dropped without waiting. Ignored fields are listed on the product's edit form, and removing one there lets the question come back. Configuration Managers can now add fields to product UI extensions in the designer, but cannot edit their HTML, JavaScript or CSS.

See Important Messages for what this means for integrations that already send custom field values.

Agent Designer: Scope on Every Agent Page

Every agent page in Agent Designer now has a "Scope" section. You choose the teams or services the agent runs on there. The Intake, Triage, Closure and Resolution Agents list the account's teams. The Knowledge and Problem Management Agents list its services. Previously, a team-scoped agent was switched on by adding it as a member of each team. Adding an agent as a team member still works, and existing settings are honored.

Shows the new "Scope" section on the Triage Agent page, where the teams the agent runs on are selected.

The section shows six entries, with the ones the agent already runs on first, and a "Show all teams" or "Show all services" link for the rest. You can filter the list and use "Select all", which applies to the entries that match the filter. A ticked entry stays selected when it is filtered or collapsed out of view. When nothing is chosen, the section warns that the agent does nothing yet. Account Administrators and Account Designers can change the scope; Auditors see it read-only.

Teams and services also show their agents. The Team page has a new "Agents" section below "Members", and "Members" now lists people only. The Service page has an "Agents" section below "Responsibilities". Changing an agent's scope in one place updates the other, and changing it on the agent page never adds or removes any of the team's people. The API is unchanged: a team's members still include its agents.

iPaaS: Connector Updates

Apply connector updates when iPaaS shows them. Xurrent updates iPaaS connectors on a regular basis to fix bugs and improve performance. This update does not list connector changes that have no visible effect. When an updated connector version is available, iPaaS shows a notification in each solution that uses the connector. Apply the update when the notification appears. The new version then goes out with the next promotion of the solution.