llm-user-testing

LLM-driven user testing simulates realistic user actions against your UI to validate flows end-to-end. Initialize artifacts with /llm-user init, then run /llm-user test <url> to execute tests and check results with /llm-user status. The Skill generates scenarios, records outcomes, and produces structured reports for QA review.

Scene-grounded validation ensures test outputs reference only UI elements actually visible in the current screen state, eliminating false positives from hallucinated interactions. This prevents tests from claiming success on actions that cannot occur, making automated testing more reliable for web and mobile applications.

Yes. This Skill applies LLM-driven testing protocols to both web and mobile UIs, including onboarding flows, dashboards, forms, and learning apps. It covers the full testing lifecycle: init, test, status, fix, and refresh commands, with structured artifact recording for standardized review.

LLM-based user simulation works across multi-step flows such as onboarding sequences, form submissions, dashboard interactions, and learning app navigation. The Skill generates domain-specific test scenarios, applies structured scoring, and records results, making it suitable for QA across diverse UI types.

Run /llm-user test <url> to execute tests, then /llm-user status to retrieve results. The Skill produces standardized reports, personas, and test artifacts in structured formats, enabling easy review and downstream analysis without manual report compilation.