Tag
12 articles tagged #Testing.
P.01Integration tests that boot every dependency are slow and flaky. Contract testing checks that consumer and provider agree on a shape without either running.
P.02Unit tests don't work on a feature that answers differently every time. Evals do. How to build a practical eval harness for an LLM feature, with real code.
P.03Full coverage means every line ran, not that your tests would notice it breaking. How mutation testing answers that, with Stryker and mutmut.
P.04QA is risk thinking, not script writing. What a strong QA engineer brings, what to test in the interview, and how to avoid hiring a regression-suite runner.
P.05Comparing LangSmith, Braintrust, and W&B Weave for LLM evaluation: what each does well, where each breaks down, and a minimum viable eval pipeline.
P.06AI tools write code fast. TDD asks you to slow down and write tests first. These two impulses seem to be in tension. Here's how they actually work together.
P.07A Playwright E2E suite that scales reuses auth state across tests, follows a folder structure built for growth, and configures CI to isolate real flakiness.
P.08k6 is a load testing tool with JavaScript scripting and CI integration. Writing meaningful tests, reading the results, and catching regressions early.
P.09Vitest runs faster than Jest, handles ESM and TypeScript natively, and shares Jest's API. The case for switching in a Vite project, and how to do it.
P.10Playwright E2E tests break because of fragile selectors, shared test state, or overly broad scope, not the tool itself. Fix those and tests survive UI changes.
P.11Unit tests confirm your code runs. They don't confirm your AI feature gives good answers. Here's how to build an eval pipeline that catches real failures.
P.12Traditional test suites break on non-deterministic AI outputs. Here's how we QA LLM features: constraint checks, golden datasets, and LLM-as-judge evaluation.