pest

Why we included this project

Teams that ship PHP applications, especially Laravel codebases, will find Pest a pleasant way to cover the usual unit and feature tests plus browser behavior without bolting several tools together. The latest major release adds browser testing on top of Playwright and an eval workflow where LLM prompts are asserted like ordinary tests, with AI scorers grading relevance or safety when an exact answer is too loose. There is also a one-shot CLI probe that a coding agent can run after an edit to confirm its change actually works, returning a definitive pass or fail instead of a guess. If you want AI-related assertions and agent verification to live inside your normal PHP test suite, this is worth a look.