Changelog

Follow new updates and improvements to kapptivate.

September 22nd, 2026

Find your elements faster with the smartphone recorder, and alerts that land in Slack, Teams & Jira

The smartphone recorder gets a major upgrade: instead of navigating through an XML tree, you now select elements directly on the screenshot. Plus, Slack, Microsoft Teams and Jira are now connected once and picked from a list when setting up alerts.

📱 Improved: Find an element, on the smartphone recorder

The recorder's "XML tree" is now Find an element. Instead of navigating through an XML hierarchy, you now work directly on the screenshot: hover to highlight, click to select.

How it works:

Three scrollable columns: hover over the screenshot to highlight elements, click to select, and the hierarchy unfolds automatically
Selectors are ranked by reliability and shown on the screenshot, with the best one preselected and marked as recommended
When multiple elements match the same selector, you're asked which one you meant, with each match numbered on screen
A search bar filters the tree by label, text or id

Coords mode: When no selector can reach your target, switch to coordinates mode. A crosshair appears on the screenshot with live x/y values you can edit. The system clearly states that this approach depends on screen resolution.

WebView support: When the element tree stops at a WebView, a banner explains why and helps you access its contents.

More steps take a locator: The element selection modal now offers Tap, Long press, Fill input, Verify element and Extract value.

API Call step is now available in smartphone tests.

Learn more about mobile recorder

🔌 New: Integration Center

Connect your tools once, use them everywhere.

Slack, Microsoft Teams and Jira are now connected through OAuth from the new Integration Center. When setting up an alert, you simply pick a channel from a dropdown instead of copy-pasting webhook URLs.

Jira creates tickets automatically. Connect a Jira project, and tickets are created automatically when an alert fires.

Track notification delivery. The Incidents table now shows the notification status for each incident, with a failure indicator and a direct link to the Jira ticket when applicable.

⚠️ Action required: The old Slack and Teams webhook fields are gone. Reconnect your workspaces from the Integration Center to keep your existing alerts working.

🧩 Improved: Custom step editor, with AI

Custom step code now opens in a fullscreen editor for better readability.

AI assistant: Describe what you want in plain language, and the AI generates the code using your test's actual selectors
Insert variable button: Browse your product's variables and insert them with the correct syntax
Library indicator: Shows which library script your step is using
Dependency warning: If your script relies on data from a previous step, you'll be warned that it won't work standalone

☁️ Improved: BrowserStack

BrowserStack is now a proper cloud agent with its own dedicated page.

See your imported devices and account details
Your API key is validated before being saved
Removing a device removes it from all agents sharing that account

✨ Other Improvements

Deep links to test editor: Share a link that opens exactly the panel, step and tab you were looking at
Pick JSON/XML attributes visually: Instead of typing paths manually, pick them from the step's last response
Smarter warning defaults: Adding a warning on an API call step starts with Response time preselected, so you only need to set the threshold
Device status explained: Hover over a device status to see what it means

🐛 Fixes

Secret variables are now masked everywhere results are shared outside the platform
Saving a test no longer clears its input variable values
Custom library steps stay within their product; exported tests keep their code
Global storage can write variables again when their group has a single variant
Slack alert failure reasons stay clickable regardless of error message content
Re-running a test works for legacy tests and uses the device from the original run
The test editor no longer crashes when reloading after a run or when showing in-test variable fields
Using the browser back button from the test editor no longer shows an error page
Save & run on a test that other tests depend on now opens run confirmation after the update
Executions missing from observability store now fall back to the legacy store
Execution steps no longer error when an API action hasn't produced a result yet
Screenshot preview labels, navigation arrows and close button work correctly
Step type picker stays on screen and scrolls properly when expanded
The "step added" confirmation no longer covers the Add step button
Hidden in-test variables appear read-only when the form is locked
Agent actions respect permissions again
Device search finds options like "iPhone"
Dialogs have accessible names and copy buttons are keyboard-navigable

Write a comment...

September 9th, 2026

Checks redesigned, in-test variables & campaign monitors 🧬

KTM 8.93

A major release. Checks are now tailored to each step type with clear plain-language conditions. Variables become a real input/output system with runtime capture and updates. Monitors can now watch entire test campaigns. And yes, dark mode is here 🤩 .

✅ New: Checks, redesigned from the ground up

Checks are no longer one-size-fits-all. Each step type now offers checks that make sense for what it does.

What's changed:

Step-specific checks: API, email and web steps each get their own relevant checks
Success and warnings are independent: A warning no longer decides whether a step passes. It's always reported separately.
Plain language conditions: "Passes if Status code = 200", "Warns if Response time > 10s"
AND/OR logic on both success conditions and warnings

Existing tests keep their exact pass/fail logic unchanged.

🔍 New: API, Email & PDF Checks

API responses: Target a specific JSON field, use variables in header names or XML paths, compare body structure with the new matches-JSON operator.

Get mail steps: Check sender, recipient, CC, subject, content and attachment name.

New Check PDF step: Download a PDF from a URL and check its extracted text, page count, file size and file name.

🧬 New: In-test variables, inputs and outputs

Variables become a proper input/output system.

Declare inputs where the test lives. The Environment tab is now grouped into sections with + buttons to create in-test variables, global variables or variable groups without leaving the form. Duplicate names are blocked with an inline error.

Capture outputs from the run. An API Call step can write its response into a variable: from a JSON path, an XML path, the entire body or a response header. Optional default values prevent step failures when nothing is extracted. Steps that create variables show them as a blue chip on the canvas.

Update a global variable at runtime. The new Update Variable step (web and smartphone) reassigns an existing variable while the test runs. A token captured at step 3 can be written into a global variable and used by whatever comes next.

Set Local Variable owns its value. The former Set Variable step now holds both name and value, accepting variables and random functions. Its variables can no longer be overridden at launch: the step is the single source of truth.

Hide sensitive values. In-test variables and Set Local Variable steps can be masked in execution results. The flag follows into run confirmation, monitors and campaign overrides. Reveal on demand when needed.

Scoped suggestions. Variables created by a step are only suggested in steps that follow it. References above their creating step show as a red tag.

Cascading renames. Renaming a variable updates checks, campaigns and monitors that use the test, with a toast reporting what was updated.

🎯 New: Monitor a whole test campaign

Create a monitor on a test campaign instead of one monitor per test.

Campaign monitors support per-test overrides and show up in the monitors list with a type filter. In the Realtime Status dashboard, drilling into a campaign shows an aggregated verdict, success count and per-test breakdown.

🧩 New: Custom step library

JS snippets can now live in a library and be referenced by your tests instead of copied into each one.

A linked Custom step shows the library step it points to, with a read-only view of its code and a way to detach. Editing or deleting a library script tells you how many tests it will update.

Custom step code is now edited directly in the step panel, like every other step type.

🌓 New: Dark mode

Pick your display mode in My Account → Preferences: Light, Dark, or Auto to follow your system.

TV mode gets its own day/night toggle in its top bar, so a screen left running in a control room can be set independently from your personal preference.

✨ Other Improvements

BrowserStack support: Back-end ready, front-end with agent setup coming soon
Send results by email: Export execution results directly from the execution details page
USSD Live export: Export a USSD Live session as a real Excel test report
Product switcher: Switch between products from the Status Details breadcrumb
Chronological order: New timeline option with oldest bucket first
TV mode timeline: Compacts itself, filter popover can select all tests at once
Cleaner step display: Extract value, SMS code, mail code, email link and Get text steps show selector and variable name on one line
Legacy Overview removed: Bookmarked URLs redirect to Realtime Status
Compact check rows: More space-efficient layout
Bulk delete checks: Right-click the "N checks" chip to delete all checks at once

🐛 Fixes

Campaign monitors now carry every in-test variable of every test, so steps get values shown in the form instead of empty values
Running a campaign with the same test multiple times keeps the variant configured for each occurrence
Typing in the middle of a variable field no longer moves cursor to the end
Opening tests created before JSON path checks update no longer crashes the editor
Email checks on Get mail steps survive every save path
Step details panel scrolls when there are more checks than fit on screen
Long JSON paths and values no longer overflow check rows
Variables extracted by a step no longer ask for an input value
"Add tests" modal in campaigns no longer comes up empty for operators with 100+ products
Reusable components no longer leak into test list screens
Web recorder can start when starting URL contains variables
Test campaigns show the selected value of choice variables, no longer showing secret variables as empty in red
Opening fullscreen images from execution results displays above the modal

Write a comment...

August 12th, 2026

Goodbye Overview, hello Realtime 👋

Overview showed the same thing as Realtime status, so the two have been merged into one. The detail view behind it has not gone anywhere either: Status details has been rebuilt as Realtime details, with a Display menu that holds everything about how the timeline is drawn, from orientation and density down to a toggle for every element on the page. This release also brings a new API Call step for web tests and makes test variables editable and overridable everywhere.

🧹 Removed: the Overview page

Overview no longer appears in the sidebar. It answered the same question as Realtime status, from the same monitoring results, with two fixed readings and a filter on zones. Anything you did there, Realtime status already did with far more control, so we kept the one that could grow and put the effort into it and into the detail page behind it.

Nothing was dropped, and nothing is broken: old Overview links redirect to Realtime status. The Overview section of the sidebar is still there, and still holds Realtime status, Incidents and Analytics.

👉 Where did the Overview page go?

📺 New: Realtime details

A row of product cards in Realtime status. The name of the first card is ringed and a pointer rests on it, marking it as the link that opens Realtime details

Realtime status cards now open the rebuilt detail page. The timeline is the same idea as before, one cell per monitor per period, rebuilt so you can shape it.

The Display menu

One button now holds everything about how the timeline is drawn.

Orientation: the vertical timeline is now the default, with each monitor in a column and time running down. The horizontal one is still there.
Density, on the vertical timeline: Normal or Compact. Compact narrows the monitor columns so many more monitors fit on screen.
Summary cards can be hidden, which gives the timeline the full height.
Display labels brings back the test type inside each bucket.
The Monitor ID, Location and 24h trend options now apply to the vertical timeline too.

Read it more easily

Filter by test, with a search field to find a test by keyword
The Monitors column is resizable, so long names stay readable
Hovering a monitor column shows its full name, and cell tooltips name the monitor they belong to
The tag row disappears on its own when none of the visible monitors is tagged
The page scrolls as one, with the table headers pinned so they stay visible as you go down

TV mode, side by side
When it is on, it hides the left navigation entirely to give the timeline the full window width.

TV mode on Realtime details: a dark full-width display with the current time and date, the summary cards, and the monitor timeline below

Your view follows you

Your settings are remembered between visits, and the URL now carries the view. A link reopens the page identically for whoever you share it with: same filters, same orientation, same options.

Faster on large products

The page renders noticeably faster on products carrying many monitors. The 24h trend column no longer redraws itself while the page sits open, which was the most visible flaw on a screen left running all day. And on long time ranges, the finest aggregations are no longer offered, since they produced tables too dense to read.

📌 The old detail view goes away next release. You can still switch back to it from Realtime details, but that switch is on its way out. Take the time to set up your view now, and tell us if anything you relied on is missing.

🔌 New: API Call step

A new step type for web tests (new experience, soon for smartphone tests) to make HTTP requests as part of your journeys.

⚠️ Update your robots first. On an older robot, the step fails. Check your robot versions from the Agents page and update them before adding API Call steps to your tests.

Build the request: method, URL, headers, query parameters and body. Every field is variable aware, so you can reuse globals, variable groups and test variables.

Assert on the response from the Checks tab: status code, timing, headers and body.

Inspect the result: the full response (body, headers, performance) is available in the test results, and the last captured response is previewed directly in the editor right after a run.

Advanced options:

Override DNS with one or several DNS servers, with templating supported in the field
"Accept insecure certificates", consistent with the rest of the web test options

🧬 New: test variables, editable and overridable everywhere

Test variables of Web and Smartphone V2 tests are no longer read only.

Edit them from the Environment tab and the step panel, always in sync with their set variable step
Create an in test variable straight from the suggestions dropdown in the test form, and give it a rich value (globals, random functions, other test variables)
Override them at launch, from the run confirmation drawer and from the monitor form, with support for variable references and random functions
Override them from a test suite: edit them in the suite's test drawer with the rich variable table, see them as tags on the test rows, and have them reused when the suite runs
See them in results: the monitor result modal now shows the variables tab, like the execution result page, and rich values are rendered as tags

✨ Other improvements

Step actions (split from here, duplicate, delete) are grouped in a dropdown menu in the test editor's right panel
Unsaved changes dialog now offers a close cross and a "Save and leave" action next to "Leave without saving"
Opening a test now expands the first editable step group instead of a leading reusable component
Clearer page titles: execution and monitor result tabs show "#id name", and alert / mail report edit pages show the entity name
Kubernetes and Usage page headers now align on the design system
Variable name inputs (set variable, extract, sms code, mail code, mail link) no longer show the dead {} suggestions button
Variables assigned by a custom step's script (test.variables.set) no longer appear in the test variables tables, since their runtime value would be overwritten by the script

🐛 Fixes

Switching between the two versions of the detail page keeps the selected aggregation and zones, not only the time range
Opening Realtime details from a card grouped by test now lands on a test filter that can be changed or cleared
A monitor with no location no longer shows an empty location pin on the table
Picking a time range now applies that range's default aggregation
Timeline tooltips of the Realtime status view are no longer cut off near the screen edges
No more double scrollbar on the timeline
Variable overrides at run confirmation: select all then typing now replaces the whole value, function tokens included
Built in random functions can be inserted in a test variable value from the run queue
The variables tab of a test result no longer crashes when a variable value is a rich template
The variable edit modal now shows the variable's description
Variables inside a variable group can now be edited by clicking their name, like global variables
Editing a variable value in the Environment tab now turns the Run button into "Save and run"
Variable creation restored from the Configurations page and the test form
API Call: checking "Override DNS" no longer crashes the form, the HTTP method is saved even when left on GET, a HEAD request no longer shows an empty response body, and the sent request headers are accessible on a GET
API test results are easier to read: assertion rows are compact and nested under their step, and an expanded step no longer leaves a large blank area
The Pretty / Raw toggle of the API response viewer is always shown and works
Times displayed next to a date range picker now follow your 12h / 24h preference
The execution mode is preserved when using "Run again" on an execution
Switching product while on a test detail page no longer leaves a blank page, the app lands on the tests list of the new product
Fixed an infinite loading screen when going back to the tests list from the legacy test edit page
The XML tree inspector button in the smartphone recorder and the copy xpath button inside the modal display their icons again
The confirmation input is now focused when opening the product deletion dialog
The app automatically reloads again when a new version is deployed
Visual polish across the V2 test editor: step cards, variable inputs, run confirmation and campaigns table

Write a comment...

June 15th, 2026

New Run Confirmation drawer & Execution Modes: run tests your way 🚀

The entire test launching experience has been redesigned. The run queue and run confirmation are now merged into a single drawer that keeps you in context. Plus, you can now choose how much data to collect on each run.

⚡ New: Execution Modes

Choose how much data to collect on each run.

Three modes available:

Fast: Disables screenshots and logs. Maximum speed. Ideal for production monitoring where you just need pass/fail.
Detailed: Keeps all context for debugging. Screenshots, logs, everything.
Standard: Middle ground. Default option.

Same test, three different durations. Pick the right trade-off between speed and observability.

Set the mode at:

Run confirmation drawer
Monitor scheduling
V2 test settings

🎯 New: Run Confirmation Drawer

The run confirmation page as you knew it is gone.

Everything now happens in a drawer that opens from wherever you are: tests page, executions, execution details... No more redirects, no lost context.

All existing actions preserved:

CSV import
Create test campaign from queue
Drag & drop to reorder
Variable overrides
Device selection
And more

The drawer shows resolved device names with availability status, an instructional banner with queue count, and all the run options you're used to (notifications, ignore errors, replay failed only, parallel execution...).

📱 Improved: Smartphone Experience

Tip banner during recording: When a Fill input step is detected, a banner reminds you that you can type directly on your keyboard
Pause and assertion options restored in the smartphone recorder

✨ Other Improvements

Monitors list pagination: The monitors list now has pagination.
Collection action: Send all tests from a collection to the run queue in one click
Remove/duplicate from drawer: Manage tests directly from the Run Confirmation drawer
Execution parameters visible: Device, starting URL, browser preset displayed for V2 tests in the drawer
Variables in drawer: Global variables and variable groups shown for V2 tests
Device pre-fill: Monitor device pre-filled from the test's default owner
XML tree dump: Now exposed in execution artifacts for download
Realtime Status filters: Banner shows only unsaved changes
Status Details page: New header with time range picker and CSV export
Confirmation notifications: Adding items to the queue shows a notification with item name or count

🐛 Fixes

Improved loading time for executions list
Fixed artifact copy button for log, text and script artifacts
Fixed broken screenshots in scheduled mail reports
Split cells in status details grid now show each status color separately
Duplicating a step now preserves its checks
Fixed step group split adding steps to wrong group
Fixed empty "Results details" modal on Realtime Status
Choice-type variables now display saved value in monitor edit form
Display changes in realtime status views now save automatically
Fixed "Test Suites" sidebar link pointing to 404
Fixed crash when creating a monitor from a test
Fixed monitoring variables panel not restoring choice values
Replay from execution now uses the device from the previous run
"Create test suite from queue" now carries over variable and owner edits
Fixed agent form not loading automatic-update time correctly

Write a comment...

May 26th, 2026

Step-level Checks are here, plus Warning status & unlimited date exploration 🎉

A packed release this week with three major features: step-level Checks to validate conditions on any action, the new Warning status that goes beyond simple Success/Failed, and the Realtime Status that finally breaks free from the 48h limit.

✅ New: Checks on Steps

Add success conditions to any step in your test.

Click on a step, go to the Checks tab, and define your conditions. If the condition isn't met, the step fails (or warns, more on that below).

Available checks to begin with:

URL: verify the current URL contains, equals, or matches a pattern
Step duration: check if the step completed within a time threshold

Combine multiple checks with AND/OR logic. More check types coming soon.

⚠️ New: Warning Status

Your tests are no longer limited to Success or Failed.

For each check, choose what happens if the condition isn't met: Failed (red) or Warning (yellow). This gives you granularity between "something is broken" and "something is slow but still working."

This status propagates to monitors as Degraded status, visible in yellow on the Realtime Status.

Important: No checks are added automatically on existing tests. To see the Warning status in action, add checks to existing tests or create new ones.

📊 New: Realtime Status Without Time Limit

No more 48h limit.

Realtime Status now lets you go back in time with ranges up to 12 months (7d, 30d, 90d, 12 months). You can also drag a period on the chart to zoom into it.

↩️ New: Undo / Redo in Test Editor

Ctrl+Z to undo, Ctrl+Y (or Ctrl+Shift+Z) to redo.

Works on everything: structure, renames, values, hooks, adding/deleting/moving steps and step groups. Text edits are grouped into 1-second sessions to undo a full burst of typing at once.

🔑 New: Personal Access Token

New section in the Profile page to generate a personal access token.

Use it to configure the Kapptivate CLI and MCP server. One click to copy, ready to use.

🤖 New: AI Visual Assertion & Spell Check on Smartphone

Two new action types available in smartphone tests:

AI visual assertion: AI-powered visual validation
Spell check: spelling verification

✨ Improvements

Test campaigns: direct move from row action menu (no more multi-select needed), collection duplication, duplicate from multi-select bar
Smartphone test creation modal: visual redesign + Refresh button to reload the app list
App reset disabled by default on new Android tests (aligned with iOS)
Variable tags: new visual style, hover reveals variable group, references in checks visible in Environment tab
Check tags: reduced padding, mini-tag style instead of raw text
Warning labels: displayed next to warning tests in multi-test results
Persistent Checks tab: clicking another step keeps the Checks tab open
Hidden operators: "not exactly" and "does not contain" removed from checks following user feedback

🐛 Fixes

Step-level timeouts not taking effect at runtime (re-save to migrate existing values)
Live preview now follows the selected test in multi-test suite
Silent reconnection when JWT expires (no more forced redirect to /login)
New users correctly attached to default security policy
Dashboard export now respects variable selection
Delays and timeouts not working on smartphone tests (new experience)
Crash and visual issues on Realtime Status
Variant and choice value properly persisted when saving a test
CHOICE variables in monitors resolved to single value before sending to runner
Adding a variant now auto-creates cells for all existing variables
Video tab always visible with clear empty state
Breadcrumb shows "Test campaigns" instead of "TestCampaigns"
Degraded status icon fixed in monitors list
"Leave without saving" alert no longer fires after successful save
And plenty more polish (typo, alignments, visual states…)

Write a comment...

May 20th, 2026

Test Campaigns 2.0, Hook Tag & a sharper smartphone toolkit 🚀

This release rebuilds test campaigns from the ground up and sharpens the whole smartphone testing flow. Plus a wave of polish improvements and fixes across the product.

🎯 Improved: Test Campaigns 2.0

Your campaigns just got a lot more flexible.

Test campaigns now let you customize behavior per campaign and organize them like a library:

Per-campaign overrides: starting URL, browser preset and test labels, all configurable independently
Variables v3 support: pick which variant of a variable group is used per variable
Collections: group related campaigns to keep large catalogs tidy

Same tests, different runs, no duplication.

🏷️ New: Hook Tag in Test Forms

A new Hook Tag component shows which hooks are wired up to the test, in both web and smartphone forms. No more digging through menus to check what's running before and after.

📦 New: Bulk Monitor Creation for Smartphone Tests

Smartphone tests join the party. You can now bulk-create monitors from a list of smartphone tests, just like the web flow. Spin up monitoring for an entire APP V2 or Web V2 test list in one go.

🔀 New: Reorder Step Groups by Drag & Drop

Drag and drop step groups directly in the test canvas to reorganize them. Combined with the new draggable cursors, restructuring a test is now a one-gesture operation.

📱 Improved: Smartphone Recorder Polish

The recorder gets a wave of refinements that make capturing smartphone tests faster and less fiddly:

Pause mode: pause the recorder without losing your session
One-tap Fill input: tap an input field and a Fill input step is created automatically
XML tree viewer: a new button on execution results opens the XML tree of the screen
Refresh app button: dedicated button to refresh the list of installed apps
Cleaner test creation modal: the smartphone test creation modal has been reworked for clarity

✨ Improvements

Variables in execution details: input and output variables now appear directly on the execution details page
Variables in Start URL: the Start URL field of the test creation modal now supports variables
Persistent sort: your sort column and direction stick across navigation on tests, monitors, dashboards, test suites, smartphones and SIMs lists
Realtime Status time ranges: added 7 days, 30 days, 90 days and 12 months on top of the existing hourly ranges
Better drag cursors: all draggable elements now have proper cursors so it's obvious what you can move

🐛 Fixes

Mail report subscriptions no longer overwrite the creator, and the recipient list is now reliably persisted
Fixed live preview not following the selected test in multi-test suites
Fixed checkbox step overflow when the description is too long
Fixed "Select a device" error when saving a reusable component without a device
Fixed selected products being reset when adding a product to a device or agent
Fixed execution page showing all execution devices instead of only the devices used by the selected test
Fixed browser preset not displayed on the monitor devices card for smartphone tests
Fixed step-level timeout defaulting to 60s instead of the test-level Selector timeout
Blocked copy to another product for tests containing reusable components
Fixed the "last update" column of the dashboards list to display the actual update date
Drag handle icon color is now consistent with the design system
Recipient tags in the mail report panel now show the email even when the recipient is not in the operator's user list
Fixed execution status stuck on "queued" in the test form after a run
Fixed crash when creating a global variable of type "Choice"
Restoring a test version now warns about variables that no longer exist
Fixed opening monitoring details from the realtime status card
Fixed unexpected logout caused by missing refresh token fallback and stale auth cookies
Fixed the "Create Monitor" button on the Realtime Status empty state
Fixed crash on the execution details page of a monitoring with multiple owners
Fixed smartphone V2 test results showing only the first step
Fixed empty owner regularly displayed on the test creation form
Fixed crash on the execution result page when there is no triggering user
The artifacts button now shows on execution steps when only systemd logs are present, keeping "Systemd output" and "Systemd error" tabs reachable
Fixed version history preview not loading anything for smartphone tests
Fixed app_reset checkbox incorrectly checked for old tests missing the key
Fixed webhook saving failure in notifications settings

Write a comment...

April 26th, 2026

New Running Mode: watch your tests execute in real time ▶️

This release introduces a completely new test execution experience. View the browser or smartphone screen live, without leaving the execution page.

▶️ New: Running Mode

A new way to follow your web and smartphone tests.

Live Preview: The browser or smartphone screen displays directly in the execution view. No need to go back to the test editor.

New UI: Step groups displayed in real time with a redesigned collapse. Progressive step-by-step display coming in a future release.

Improved Navigate to URL: This step group now shows details directly in the collapse. Perfect for validating loading times.

Smooth Transitions: Cleaner transition from "running" to "success/failed" status.

Edit Related Test: Access the test editor directly from running mode.

✨ Improvements

Insert Step button: Add a step between two existing steps directly in the canvas
Cancel Execution on Test Suites: Cancel a running test suite
Skipped step groups hidden: Execution results no longer show noise from skipped groups
Assertion mode auto-disabled: After a click in the smartphone recorder, assertion mode turns off automatically
More accurate execution times: Failed tests are excluded from weekly averages
Aligned timers: Step group timers aligned with step timers inside the collapse
Scroll to top: Automatic scroll to top on pagination change
Password managers blocked: No more autofill on shared dashboard password fields

🐛 Fixes

Fixed Realtime Status count display (time filter aligned with GROUP BY intervals)
Fixed day selection alignment in date picker
Agents list now sorted in natural/human order
Fixed smartphone test crash when Environment tab is focused
Fixed Replay button not working on results page
Fixed execution creator not displaying on execution details page
Fixed shifting and unaligned step group collapse layout
Fixed monitoring list counter to match filtered list
Fixed background execution list not filling available space
Display global variables dropdown on shared dashboards
Prevent users from granting permissions they don't have when creating custom roles
Fixed inconsistent device information on execution details page
Fixed device selection not applied when creating monitor from execution parameters modal
Fixed actions column width in browser presets and tests tables
Fixed status colors on public dashboards
Fixed inability to scroll multi-test result details during execution
Fixed incorrect French translation for maintenance exit button
Fixed CSS priority issue via lazy loading of monitoring and supervision routes
Test editor and monitoring pages now reflect renamed variables immediately

Write a comment...

April 21st, 2026

Variables Variants are here: general variables, variants & smarter composition ✨

This release fundamentally changes how you manage data across your tests. Define variables once at the workspace level, create variants for different scenarios, and compose your test executions with the right data in just a few clicks.

📦 New: Variables V3

The new version of variables is now live across all environments.

General Variables: Define a variable once, use it everywhere. A single value shared across all your tests, perfect for environment URLs, API endpoints, or any constant.

Variable Groups with Variants: Variable groups now support multiple sets of values. Think user profiles: create a "Profiles" group with variants like "Admin", "Standard User", "Guest". Each variant holds its own values (login, password, permissions...).

Easy Composition: When executing a test or configuring a monitor, select which variants to use. Mix and match to create the exact scenario you need. No duplication, no manual overrides.

🔄 New: Replace Test

Swap an existing test with a new one while keeping all connected monitors, historical data, and reporting continuity intact. No more recreating monitors when a test needs a major overhaul.

🌐 New: Browser Preset at Monitor Creation

Select a Browser Preset directly when creating a monitor. Your test runs under the exact same conditions as manual execution. A default preset is now automatically applied for web tests.

🔐 New: Accept Insecure Certificates

Testing on staging with self-signed certificates? The new web testing experience now includes an "Accept insecure certificates" option.

✨ Improvements

Agent settings: New "About" panel to edit agent names
Dismissable banners: Communication banners now have a persistent close button
Keyboard navigation: Escape key closes selection banners on all list pages
Validation guardrails: Can't save a monitor or test without required device and browser preset
Unresolved variables visible: Test variables that aren't resolved now show clearly in the monitoring form

🐛 Fixes

Fixed live browser disconnecting after ~90 seconds
Fixed public shared dashboards failing to load charts
Fixed crash when switching products on the variables page
Fixed monitor save losing variable actions
Fixed incidents page crash
Fixed selection banner not showing in Reusables
Fixed banner displacing the action bar in list pages
Fixed capacity tab layout in web agent details
Fixed breadcrumb collection names
Blocked access to archived workspaces for non-super-admin users
Added all status options to monitor filtering

Write a comment...

March 10th, 2026

Error exclusions, public dashboards & a cleaner navigation ✨

This release brings powerful new ways to reduce noise and share insights, plus a major reorganization of how you access agents and devices. Less clutter, more control.

🔕 New: Error Exclusions for Alerts

Not every error deserves to wake someone up at 3am.

The new Error Exclusions feature lets you define patterns to filter out expected or harmless errors from triggering alerts. Find it under Incidents → Configuration → Error exclusions.

Create rules based on error patterns (e.g., "robot unreachable" during planned maintenance) and focus your alerts on what actually matters.

🔗 New: Public Dashboard Sharing

Need to share a dashboard with someone outside your workspace? Now you can.

Generate a password-protected public link for any dashboard. Recipients can view the dashboard without needing a Kapptivate account.

Password protection required (minimum 8 characters)
Share via link or email
Revoke access anytime

Perfect for sharing status pages with clients or stakeholders.

🏠 Redesigned: Agents & Devices in Main Navigation

We've moved Devices lab, Browser presets, and Agents out of the admin panel and into the main sidebar under Equipments.

What this means:

All users can now see and browse agents, smartphones, and devices
Actions are permission-based: only users with the right roles can perform maintenance, reboot, or configuration changes
Cleaner separation between viewing and managing

No more hunting through admin menus to check device status.

⚙️ Redesigned: Settings & Account Management

We've reorganized the Settings panel for clarity:

My Account: Profile & Preferences in one place
Workspace: Products, Users, Teams, Roles
Administration: Notifications, API Keys, Security policies, Event log

A simpler structure that's ready for future enhancements.

✨ Improvements

Smoother test creation: Improvements to both smartphone and web test creation flows
Custom capabilities restored: Custom browser capabilities are accessible to all users again (sorry for the temporary restriction!)

Another step toward a cleaner, more intuitive Kapptivate.

Let us know what you think! 💬

Write a comment...

February 25th, 2026

Small details, big impact: smoother navigation and smarter test controls 🎯

This release focuses on the little things that make your daily testing workflow more intuitive. Better navigation, more control over timing, and clearer execution details.

⏱️ New: Step Delay & Selector Timeout Controls

Fine-tune your test timing with new controls for web tests.

You can now configure step delays and selector timeouts directly in your test settings. Perfect for handling slow-loading elements or adding breathing room between actions.

🔗 New: Intent Field for Get Email Link

The Get email link action just got smarter.

The new Intent field lets you specify exactly which link to retrieve from an email. Whether it's "the first link in the text," a password reset link, or a verification URL, you're now in control of what gets captured thanks to our AI.

🧭 Improved: Collection Navigation

Moving through your test collections is now more intuitive.

We've reworked the back buttons and breadcrumbs to make navigation feel natural. No more wondering where you'll end up when you click back.

📋 Improved: Browser Preset in Execution Details

You can now see which Browser Preset was used directly in the execution details page.

At a glance, know if that test ran on Chrome Desktop, mobile Safari, or any other configuration. No more guessing.

✨ UX Polish

Preserved element selection: When changing an action type, your recorded element stays selected
Consistent iconography: Updated icons across the interface for better visual coherence

Write a comment...