Back to marketplace

Playwright

Gives Claude autonomous headless browser control — it spins up a browser, clicks through your UI, checks everything works, and fixes broken tests.

Development / Testingdevelopmenttesting
by AgentVoltv1.0.0Published 1y ago
Free to sign up · every skill included with AgentVolt Pro

Before & after

Same prompt, same agent — the only difference is the skill.

Prompt“Automate tests for our checkout flow”
Beforegeneric AI agent
Testing checkout end-to-end is really important but also complex—lots of edge cases and payment flows to handle. You could use Playwright, Cypress, or Selenium depending on your stack. Do you have a testing framework already? What's your tech stack—React, Vue, something else? Are you testing against a staging environment or production-like setup? Once I know more, I can help you design the test strategy.
Afterwith Playwright
PLAYWRIGHT TEST SUITE: Checkout Flow

TEST COVERAGE (14 scenarios)
✓ Happy path: Add item → enter shipping → enter payment → confirm
✓ Promo code: Apply 20% off, verify total updates
✓ Cart persistence: Close browser, reopen, cart intact
✓ Validation: Missing email → error shown, submit blocked
✓ Payment: Visa, Amex, Apple Pay (3 methods)
✓ Shipping: Standard vs Express, cost updates
✓ Region: US, EU (VAT applied), CA (GST applied)
✓ Inventory: Item sells out mid-flow → flash error
✓ Mobile: Responsive layout on 375x812 viewport
✓ Accessibility: Form labels present, ARIA roles correct
✓ Performance: Checkout loads <2s, no Core Web Vitals breach
✓ Errors: Network 500, payment decline, retry logic
✓ Security: No XSS (payment form isolated), CSRF token present
✓ Race: Simultaneous add to cart from two tabs

TEST COMMAND
Playwright ran 14 tests across Chromium, Firefox, Safari
Result: All green
Execution time: 47 seconds
Last run: 2 hours ago (auto re-run on main branch)

IF FAILURES DETECTED
→ Playwright screenshots diff before/after
→ Video recording of failed step (available in CI artifact)
→ Retry once (network flake handling)
→ If still failing: block merge until fixed

CI INTEGRATION
GitHub Actions: test on push to main + PRs
Block merge: any test failure
Notification: Slack to eng-team on failure

About this skill

name: playwright description: Use when Gives Claude autonomous headless browser control — it spins up a browser, clicks through your UI, checks everything works, and fixes broken tests.

Playwright

The Playwright skill gives Claude full autonomous browser control. It spins up a headless browser, clicks through your app like a real user, validates every feature works, and fixes tests when they break. Even Anthropic's Boris Churnney built a custom /go slash command on top of Playwright to autonomously test his own code.

What you get

  • Autonomous browser testing, UI click-through validation, automatic test fixing when tests break, headless browser control across Chromium, Firefox, and Safari.

Customize your output

  • Configure which browsers to test (Chromium, Firefox, Safari) and define test coverage depth and which flows to prioritize.

Example output

Playwright ran 14 UI tests, caught broken checkout flow on step 3, auto-fixed the selector, confirmed passing. All green.

Best for

Developers who hate writing tests, vibe coders shipping fast, QA-conscious solo builders.

SKILL.md preview

SKILL.md
---
name: playwright
description: Use this skill to give Claude autonomous headless browser control for clicking through a UI, validating features, and fixing broken tests.
version: 1.0.0
category: Development / Testing
author: AgentVolt
license: proprietary
tags:
  - development
  - testing
  - standard
---

# Playwright

Drives a headless browser to exercise an application like a real user, verifying features work end-to-end and repairing tests when the UI changes underneath them.

## When to use

… (sign up to view the full skill)
Sign up to view, copy, and install the full skill