Playwright MCP

What it does

Playwright MCP gives an agent browser tools: open a page, click, type, read, and check results. It works on Playwright's accessibility tree, so the model reads structured page data instead of guessing from pixels.

Why it is useful

• Fast and light: no vision model is needed to use a web page.
• Good for testing your own web app, reproducing bugs, and filling repetitive web forms.
• One npx command works in Claude Code, Codex, Cursor, VS Code and most other MCP clients.

How to use it

claude mcp add playwright npx @playwright/mcp@latest

Then ask the agent to "open localhost:3000 and check that the sign-up form works".

Watch out for

• For plain coding agents, Microsoft now suggests its Playwright CLI with skills, because CLI calls use fewer tokens than large MCP tool schemas.
• MCP is still the better fit for long, exploratory browser sessions that need state.