playwright-skill MCP — Use Cases, Install & Live Demo

Why use it

Key features

Model-invoked — Claude writes custom Playwright code tailored to your ask
Headed browser by default so you can see the automation run
Built-in run.js handles module resolution so scripts just work
On-demand Playwright API reference loading to keep the context lean
Handles testing, form filling, click-throughs, broken-link checks, responsive checks

Live Demo

What it looks like in practice

playwright-skill.replay ▶ ready

0/0

Install

Pick your client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "playwright-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/lackeyjb/playwright-skill",
        "~/.claude/skills/playwright-skill"
      ],
      "_inferred": true
    }
  }
}

Open Claude Desktop → Settings → Developer → Edit Config. Restart after saving.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "playwright-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/lackeyjb/playwright-skill",
        "~/.claude/skills/playwright-skill"
      ],
      "_inferred": true
    }
  }
}

Cursor uses the same mcpServers schema as Claude Desktop. Project config wins over global.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "playwright-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/lackeyjb/playwright-skill",
        "~/.claude/skills/playwright-skill"
      ],
      "_inferred": true
    }
  }
}

Click the MCP Servers icon in the Cline sidebar, then "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "playwright-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/lackeyjb/playwright-skill",
        "~/.claude/skills/playwright-skill"
      ],
      "_inferred": true
    }
  }
}

Same shape as Claude Desktop. Restart Windsurf to pick up changes.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "playwright-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/lackeyjb/playwright-skill",
        "~/.claude/skills/playwright-skill"
      ]
    }
  ]
}

Continue uses an array of server objects rather than a map.

~/.config/zed/settings.json

{
  "context_servers": {
    "playwright-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/lackeyjb/playwright-skill",
          "~/.claude/skills/playwright-skill"
        ]
      }
    }
  }
}

Add to context_servers. Zed hot-reloads on save.

claude mcp add playwright-skill -- git clone https://github.com/lackeyjb/playwright-skill ~/.claude/skills/playwright-skill

One-liner. Verify with claude mcp list. Remove with claude mcp remove.

Use Cases

Real-world ways to use playwright-skill

Write a smoke test for a signup flow on demand

👤 Developers who want a test without hand-writing Playwright ⏱ ~10 min beginner

When to use: You pushed a new signup flow and want to know it still works end-to-end.

Prerequisites

Node + Playwright installable — The skill will install browsers via npx playwright install if missing

Flow

Describe the flow

Use the playwright skill. Open localhost:3000, sign up with [email protected] / Pass123!, verify we land on /dashboard.✓ Copied

→ Claude writes a Playwright script, runs it headed, reports pass/fail
Debug failures interactively

It failed at the email input — add a screenshot right before the error.✓ Copied

→ Next run captures and returns the screenshot

Outcome: A one-off smoke test you can optionally save as a persistent spec.

Pitfalls

Selectors chosen are brittle (e.g. nth-child) — Ask Claude to prefer role/label selectors

Combine with: filesystem

Find all broken links on a documentation site

👤 Docs maintainers ⏱ ~15 min intermediate

When to use: You suspect link rot after a restructuring.

Flow

Describe the crawl

Use playwright to crawl https://docs.example.com up to depth 2, log every link, and return the 404s.✓ Copied

→ Claude writes a BFS crawler, runs it, outputs a list

Outcome: A triaged list of dead links to fix.

Pitfalls

Crawler hits the site hard — Cap concurrency at 2-3 and add small delays

Spot-check responsive rendering across 3 viewports

👤 Frontend devs validating a layout change ⏱ ~5 min beginner

When to use: You changed the header and want to see how it renders on desktop, tablet, and mobile.

Flow

Request screenshots

Use playwright to screenshot localhost:3000 at 1440, 768, and 390 width. Save them.✓ Copied

→ 3 PNGs saved and paths returned

Outcome: Three screenshots for quick visual review.

Pitfalls

Screenshots don't wait for fonts/images — Add a networkidle wait before snapping

Combine with: filesystem

Combinations

Pair with other MCPs for X10 leverage

playwright-skill + filesystem

Save test specs and screenshots alongside the code they test

Write the smoke test and save it under tests/e2e/signup.spec.js.✓ Copied

playwright-skill + github

Run the test, fail the workflow, post results on the PR

Run the playwright spec against the deploy preview, then post results as a comment on PR #42.✓ Copied

playwright-skill + ios-simulator-skill

Playwright for web, ios-simulator-skill for native — use both in a cross-platform test session

Run the playwright web flow and the ios-simulator native flow in sequence, report both.✓ Copied

Tools

What this MCP exposes

Tool	Inputs	When to call
Custom Playwright script generation	task description	Any browser automation ask
run.js executor	script path	After script generation
Headed/headless toggle	flag	CI vs local debugging

Cost & Limits

What this costs to run

API quota: None
Tokens per call: 1-5k per test run, depending on how much Playwright reference Claude pulls in
Monetary: Free — you only need Node + Playwright browsers on disk
Tip: Keep prompts focused; don't ask for 'test everything' in one shot.

Security

Permissions, secrets, blast radius

Credential storage: No credentials in the skill. If your tests hit authenticated pages, keep creds in env vars — never in the generated script.

Data egress: Only to sites your test visits

Generated scripts run with full local-process privileges — review before running against production URLs.

Troubleshooting

Common errors and fixes

'browserType.launch: Executable doesn't exist'

Run npx playwright install chromium

Verify: npx playwright --version

Selector fails intermittently

Switch to role/label selectors; add await expect(...).toBeVisible() waits.

Script hangs

Claude likely forgot to close the browser. Tell it to always wrap in try/finally with await browser.close().

Alternatives

playwright-skill vs others

Alternative	When to use it instead	Tradeoff
playwright MCP	You want structured MCP tools (click, fill, snapshot) rather than code generation	MCP abstracts away code; this skill writes real scripts you can reuse
Puppeteer skills / chrome-devtools MCP	Chrome-only workflows, DevTools protocol features	Less cross-browser

More

Resources

📖 Read the official README on GitHub

🐙 Browse open issues

🔍 Browse all 400+ MCP servers and Skills