Videos
QF-Test Blog
Manual
Tutorial
FAQ
AI agents such as Claude Code, Cursor, or GitHub Copilot have long been part of everyday work for many development teams – they write code, refactor it, and suggest solutions, often without a human having to initiate every single line. The obvious question that follows: why shouldn’t the same agents take over testing as well?
This is exactly the question QF-Test addresses – our tool for professional UI test automation. In a recent interview “How QF-Test Makes Agentic Testing Possible”, our software developer and trainer Max Melzer explained to mgm technology partners how we are integrating agentic testing into QF-Test step by step – and why this shift is more than just a new buzzword.
Why test automation is changing right now
With the rise of powerful AI coding agents, expectations are shifting: teams no longer just want to run tests automatically – they want to delegate the entire testing process, from planning to evaluation, more heavily to AI. “Our customers are incredibly interested in figuring out which parts of the testing process they can hand off to AI,” says Max Melzer. This is precisely where our staged approach to agentic testing comes in.
Three levels of agentic testing
We are deliberately developing agentic testing in QF-Test in several stages that build on one another.
Level 1 – AI as a semantic validator (already available): Some test results cannot be verified with a simple target/actual comparison, such as the responses of an integrated chatbot. QF-Test can pass such results to a language model for evaluation, which assesses whether the response is plausible in terms of content. This turns rigid string comparisons into a semantic check.
Level 2 – AI generates test suites (from QF-Test 11): Instead of recording tests manually, AI will provide an initial draft in the future – based on existing test plans or tickets, or by exploring the application on its own and identifying testable areas. The expectation here is realistic: “It’s never the case that you can just take it and say: we’re done now,” says Max Melzer. “But it takes a lot of work off your plate — especially at the start.” The first draft is the basis that experienced testers refine further – not a finished test package without human oversight.
Level 3 – QF-Test as an MCP server (the real breakthrough): With QF-Test 11, the tool gains an integrated MCP server. Via the Model Context Protocol, AI agents such as Claude Code can then access QF-Test’s capabilities directly – launching applications, running tests, checking results – without anyone having to open QF-Test manually. An experimental preview of this MCP server is already available as a public preview, and the first customers are already integrating it into their CI/CD pipelines.
What sets QF-Test apart from specialized tools
Other tools specializing in web testing now offer MCP support as well. For QF-Test, the decisive difference lies in its platform breadth: we test native Windows applications, Java UIs, Android and iOS apps, and PDF documents using the same approach as web applications. Anyone running a heterogeneous application landscape in their enterprise environment – as is usually the case – gets a single test agent for all areas instead of having to use a separate tool for each platform.
Outlook: what’s coming with QF-Test 11?
Test suite generation and the production-ready MCP server will be introduced with QF-Test 11, which is planned for 2026. Interested customers and partners can already get preview versions, so that real feedback from real-world projects can flow into further development.
One thing remains important: agentic testing does not replace experienced testers. AI-generated test suites provide a structured starting point – the professional validation and further development remain the responsibility of the QA teams.
Learn more: Read the full interview with Max Melzer, software developer and trainer at QF-Test, at mgm technology partners: “How QF-Test Makes Agentic Testing Possible”.
More details about the AI integrations in QF-Test
When tests become intelligent: AI-driven checks with QF-Test
In this special webinar, we’ll show you how to get the most out of the new AI integrations in QF-Test. With QF-Test 10, you can use AI to test non-deterministic UIs, validate UI components based on semantic criteria, generate test data, and much more.