End-to-End Testing an Astro Site Against the Real Production Build
Why I run Playwright against astro preview instead of the dev server, and the pile of gotchas—WebKit, worker limits, analytics noise—I hit getting there.
Why I run Playwright against astro preview instead of the dev server, and the pile of gotchas—WebKit, worker limits, analytics noise—I hit getting there.
The Testing Trophy, use case coverage, and mocking at the network boundary—plus the honest answer to a team coverage gate, which is not to delete it.
A test suite at 99% coverage, all green, and the main feature of the app is broken. Nobody wrote a bad test. That is the interesting part.