currents_dev

Compare Playwright MCP, the CLI, test agents and a prompt-only assistant on your own suite, then score and gate the AI Playwright tests they write.

Playwright 1.64 in CI: run --shuffle nightly and keep the seed, keep shared-account tests apart with locks, and catch the breaking changes.

Cut GitHub Actions costs for Playwright without losing coverage. Cache browser binaries, scope triggers, set job timeouts, shard correctly, track flaky tests, and tighten artifact retention.

How Sherpany reduced Playwright test flakiness from 5% to near zero, cut suite runtime by 39%, and reclaimed QA time with Currents.

A seven-check review checklist for AI-generated Playwright tests, covering hallucinated assertions, intent-weakening heals, and UI-driven state setup before they reach CI.

JavaScript test runners share status words and disagree about what the words mean. Playwright and Cypress use "skipped" for opposite cases. Jest carries 7 statuses but not "flaky". This guide translates each runner's vocabulary and shows how runners deriv

Cheaper rates, bigger volume discounts, a new self-serve Business plan, and fairer overage. Here's what changed and why.

Playwright API tests that pass locally still fail in CI? Here's the fixture architecture, per-worker isolation, and observability practices that fix that.

Not every component needs a real browser. Here's a clear decision framework for when Playwright component testing adds value and when Jest already covers it.

A Playwright custom reporter takes 20 minutes to prototype and months to maintain. Here's what it actually takes to build one that holds up in production CI.