15 min
Playwright 1.65 Quietly Changed What Agents See
Playwright 1.65 aria snapshot changes rename buttons, drop collapsed text, and block inert actions. Pin one version for tests and MCP, and diff what agents see.
Posts tagged "self-healing"
All
ai ai-agents ai-testing api-testing appium career case-study cicd claude code-review consulting contract-testing copilot cypress debugging design-patterns devops docker enterprise flaky-tests framework-design github-actions hiring java jenkins leadership lessons-learned llm localization locators loose-coupling migration parallel-execution performance-testing playwright qa-strategy reporting sdet sdet-career security selenium self-healing shift-left slack team-building test-automation test-data test-generation testing testng tight-coupling trace-viewer war-story
15 min
Playwright Healer Found the Bugs, Then Skipped the Tests
Playwright's team says you're better off without the healer agent. In my demo it found both regressions, marked them test.fixme(), and the CI run exited 0.
15 min
A Self-Healing Locator Hid a Bug for 3 Weeks
A self healing test false positive kept our suite green for 23 days while paid checked bags vanished from real bookings. The heal log, the postmortem, the fix.
12 min
Don't Use Playwright Agents Until It's Boring
When to use Playwright AI agents: not until your suite is boring. The readiness floor nobody defends, why 200 tests is the wrong gate, and how to score yours.
20 min
3 AI Testing Tools Hid the Bugs I Planted (2026)
An honest AI testing tool evaluation: three self-healing tools, one enterprise suite, six planted bugs. Heal rates hit 87% — and every tool hid real ones.
12 min
Put a Token Budget on Your Playwright AI Agents
Playwright's healer and generator agents loop on tokens. Without a per-PR AI token budget that fails the build, CI cost scales with flakiness, not value.
16 min
Run Playwright's Healer in Shadow Mode Before You Trust It
Everyone quotes a heal rate for Playwright's Healer. Nobody sourced it. Build a shadow-mode harness and measure self-healing precision on your own suite.
12 min
Don't Make an LLM Your Visual Regression Gate
A multimodal LLM is a great visual-regression triage tool but a dangerous merge gate. Here's where an LLM oracle earns its place in a Playwright suite.
12 min
Anchor AI Test Agents to the Accessibility Tree
Self-healing agents that repair Playwright tests against the DOM optimize for green, not truth. Anchor them to the accessibility tree — the contract users see.
11 min
Playwright's Healer Agent Will Hide Your Race Conditions
Playwright's Healer agent auto-fixes red tests to green. In a parallel enterprise suite, that's how real regressions ship silently. Here's the guardrail I use.
12 min
Self-Healing Locators Aren't Magic — Here's How They Work
Most self-healing test tools use simple fallback chains, not real AI. Learn the 4 approaches to AI locator repair and which one fits your framework best.