TestMu AI Launches Kane CLI Exploration, an AI Agent That Maps an Entire Web App From a Single URL and Turns It Into Tests
The exploration agent walks every page and flow of a web app on its own, flags UX, security and accessibility issues, and builds the map Kane CLI uses to write, run and ship tests to CI, even when no requirements exist
TestMu AI, the Full Stack Agentic AI Quality Engineering platform, announced Kane CLI Exploration, an autonomous agent that learns a web application before anyone writes a test. Given a single URL, it walks the entire app on its own, flags problems along the way and gives Kane CLI the context it needs to generate and run tests, even when no requirements, user stories or acceptance criteria exist.
Also Read: AiThority Interview with Tim Jurka, VP Engineering at LinkedIn
“Teams now write careful instruction files for the agent that writes their code, while the agent testing that code gets a URL and nothing else. Kane CLI Exploration builds that context from the app itself.” Mudit Singh, Co-Founder and Head of Growth, TestM
Every test plan starts from something written down, and many apps never had it. Legacy products lose their specs when the people who wrote them move on, and apps a coding agent built in a week often have no record of intended behavior beyond the prompt history. With nothing to test against, a testing agent ends up grading the app against itself.
Starting from the URL a developer provides, the exploration agent moves through every page, junction and flow it can reach, the way a new QA engineer clicks through a product in their first week. For each page it records the controls, the flows in and out, and the states the screen can be in, such as a form before and after a validation error. Each exploration runs across dozens of parallel browsers in the TestMu AI cloud, so a large app can be mapped in hours, and anyone on the team can open a live browser session to watch the agent work. Without being told what to look for, it also flags UX bugs, security issues and accessibility concerns on pages that existing Playwright, Selenium or Cypress suites never reached.
The map then becomes the testing context. Kane CLI reads it the way it would read a requirements document, writes use cases and acceptance criteria from it, runs the resulting tests in real browsers and saves an Evidence Pack for every run. A person who knows the product reviews the generated tests, and the approved ones run in CI on every future pull request.
“Teams now write careful instruction files for the agent that writes their code, while the agent testing that code gets a URL and nothing else, and in a lot of apps there is no spec left to hand it,” said Mudit Singh, Co-Founder and Head of Growth, TestMu AI. “Kane CLI Exploration builds that context from the app itself. For a team inheriting a legacy product, or shipping AI-written code faster than anyone can document it, that is the difference between a blank test file and a real starting point.”
Also Read: Architecture, Not Intelligence, Will Define the Future of AI Headphones
[To share your insights with us, please write to psen@itechseries.com]
Comments are closed.