May 2026 marks a turning point for how QA teams think about automation. This month's releases push Katalon True Platform closer to a world where you describe what you want to test, and the platform handles the rest: from generating test cases to running them, then handing you a ready-to-use script. Here is everything that shipped in May.
Run with AI: From requirement to reusable script in one session
The headline feature of May is Run with AI, a prompt-to-script agentic loop that connects every step of the manual-to-automated testing journey without handoffs.
Here is how the flow works:
- Write a test in plain language.
- Run with AI dispatches an autonomous agent to execute it against your live application, clicking, filling fields, verifying states, and recovering from unexpected UI changes on its own.
- When the agent finishes, it returns a ready-to-run Katalon Studio script built from exactly what it did, in the same session, one agent, no handoff.
The result: the manual test you wrote becomes a version-controlled, re-executable automation script with no hand-coding required.
Run with AI is currently in beta. The complete flow, including auto-save directly to your cloud repository, is coming in an upcoming release.
What else the agent handles:
- Private and internal apps. The agent can now test applications behind a firewall or on localhost via a secure tunnel, with no inbound network changes needed.
- Responsive and viewport testing. A new set_viewport action lets the agent automatically resize the browser mid-test, so a step like "View dashboard on mobile (375x667)" is handled entirely by the agent, no manual configuration required.
Learn more about Katalon True Platform • See release notes
Katalon AI Assistant and Test Generator: Smarter, more transparent, and more controllable
May brought a wide set of improvements to Katalon AI that address real friction points in daily use, from session continuity to transparency about what the AI is actually doing.
What changed:
- Conversations that remember. A persistent orchestrator means sessions survive page refreshes and long flows. Context no longer resets when you least want it to.
- Run an entire suite with one prompt. You no longer need to list individual test cases. Tell Katalon AI to run a suite and the agent executes the whole thing.
- Mid-task clarification. Katalon AI now asks when a request is ambiguous instead of guessing, which leads to better outputs and fewer do-overs.
- Visible reasoning. The chat panel now surfaces the actual tools Katalon AI calls and streams its reasoning in real time, so you can see what is happening, not just the result.
- Artifact sidebar. Every file Katalon AI generates during a conversation appears in a sidebar you can preview and download without leaving the chat.
- Multimodal input. Katalon AI now reads screenshots and uploaded documents and understands core platform concepts, giving it better context for more complex requests.
- Custom AI prompts. You can configure prompts so generated test cases match your specific application, domain, and conventions, not generic boilerplate.
- Inline review. Accept or reject AI-generated test cases as cards directly inside the chat, without navigating away.
Try Katalon AI Assistant • See release notes
True Production Insights: More control over generated tests
True Production Insights, which generates test cases from real user journeys captured in production, received several updates focused on giving teams more control over what gets generated and how those tests are maintained over time.
New in May:
See release notes
MCP and Skills: Build your own AI testing workflow
For teams building custom AI-driven testing flows, May delivered a cleaner, more reliable MCP foundation and a first look at a shared skill library.
What changed:
- Consolidated MCP tool set. Over 20 overlapping legacy tools have been retired. Any agent can now reliably create test cases, link requirements, build test suites, and run with AI in a predictable, well-documented way.
- Chained operations. Tool responses now return entity IDs, which means an agent can create a test case, link it to a requirement, and run it, all in a single uninterrupted flow.
- Katalon True Skill Library. A library of ready-made agent skills covering generation, execution, triage, and reporting is now available to browse and download. Drop any skill directly into your own flow.
- Wider repository support. ZIP-based Katalon Studio projects are now discoverable and executable through Katalon AI, expanding the range of project structures the platform can work with.
Browse the skill library
Katalon Studio 11.2.0 & 11.2.1: An AI-forward release for desktop
Studio 11.2.0 brings the AI improvements from the platform into the desktop IDE, with notable upgrades to how failures are analyzed and how broken tests recover themselves.
What is new:
-
AI Failure Analysis categories. The Katalon AI Assistant can now analyze execution reports and group failures by AI-determined root cause. Instead of scrolling through individual error messages, you see failures organized by category, making it faster to identify systemic issues versus one-off flakiness. AI-generated explanations are clearly marked with an AI tag to distinguish them from standard failure messages. This works in both HTML reports and email reports. Learn more about AI Failure Analysis
-
Smarter AI Self-Healing. When classic self-healing, which tries alternative locators when the original breaks, cannot recover a broken object, Katalon Studio now escalates to AI self-healing. The updated AI self-healing leads with the element screenshot as its primary input, giving the LLM richer visual context to identify the correct target element and propose an alternative locator. You review and approve or discard suggested locators in the Self-healing Insights panel before they become permanent. Learn more about self-healing
-
TestCloud Tunnel for mobile native apps, advanced recorder config, and a consolidated Studio MCP tool set.
- View execution results on True Platform. New links now appear directly in Job Progress and the Katalon AI Assistant chat window, letting you open your run's full results on True Platform without leaving your current context.
Download Studio 11.2.1 • See release notes
Platform and admin: quality-of-life improvements
Several smaller but meaningful updates landed across the broader platform this month.
- Bitbucket Data Center integration. Bitbucket Data Center repositories are now supported, expanding the range of source control environments that can connect to the platform.
- More reliable Xray (Jira) sync. The Xray integration for Jira is now more stable, reducing sync failures that interrupted traceability workflows.
- Requirement Traceability Report. A new report maps test cases to requirements and surfaces coverage gaps, giving QA leads and project managers a clearer picture of what is and is not tested.
- Visual Testing improvements. Unresolved visual checkpoints now surface at the top of the review queue, with easier comparison and one-click resolution.
- Guided new-trial onboarding. New accounts now see a first-action card layout that walks them into AI-assisted testing from the moment they open the platform.
- Smaller wins. Manual test runs now auto-generate a default name, and the add-cases/suites dialog has been simplified for quicker selection.
See release notes
Wrapping up
May's releases center on one idea: reduce the distance between describing a test and having a reliable, reusable automation artifact. Run with AI closes that loop end to end. Agent Mode in Katalon AI Assistant and Katalon Studio makes the AI a genuine collaborator on both the platform and the desktop. True Production Insights, MCP, and the platform improvements fill in the edges.
The beta for Run with AI is open now. Try the full flow and tell the team what needs to change.
▶️ See what shipped next in the Katalon Product Roundup June 2026.