Login/Register
Microsoft

Playwright MCP

Official

Browser automation through structured accessibility snapshots.

Playwright MCP turns browser state into structured accessibility snapshots that agents can reason over. It is useful for repeatable browser tasks, UI verification, and debugging flows where semantic page information is more reliable than coordinate-only automation.

ActivestdioAuth: None21 tools
Developer ToolsAI & AutomationProductivity
GitHubWebsite

Tools

Selected capabilities exposed to a connected AI client.

21 tools
browser_navigate

Open or move the connected browser to a target page.

browser_snapshot

Capture structured page state for inspection and follow-up actions.

browser_click

Activate a page control or other interactive element.

browser_type

Enter structured input into fields and forms.

browser_tabs

A callable capability exposed by this MCP server.

browser_take_screenshot

Capture a visual record of the current page or browser state.

Bowora highlights 6 of 21 tools. The source repository contains the complete, current tool schema.

Setup and access

Check the connection method, access level, and compatible clients before enabling the server.

npx @playwright/mcp@latest
  1. 1Install the MCP package in the environment where the client runs.
  2. 2Add the stdio command to your MCP client configuration.
  3. 3Use a dedicated browser profile or test environment before signing into sensitive services.
Transport
stdio
Connection
Local stdio MCP
Authentication
No separate MCP account authentication
Access
Can navigate, inspect, and interact with the connected browser session
Registry status
Active
Last updated
Jul 2, 2026

Compatible clients

ClaudeCodexCursorVS Code

Connects to chromium, firefox, and webkit browser sessions. Use the least-privileged credentials and access scope that support your workflow.

The agent can see pages, page content, and session state available to the connected browser. Use an isolated profile for sensitive workflows.

Availability: Open source package; browser and test-environment costs depend on your setup.

Good for

  • Test user journeys in a real browser
  • Inspect pages and interact with forms or controls
  • Automate browser-based research and verification

FAQ

What does Playwright MCP MCP do?
Playwright MCP turns browser state into structured accessibility snapshots that agents can reason over. It is useful for repeatable browser tasks, UI verification, and debugging flows where semantic page information is more reliable than coordinate-only automation.
How do I connect Playwright MCP MCP?
Add it to your MCP client over stdio using: npx @playwright/mcp@latest
Does Playwright MCP MCP require authentication?
No separate MCP account authentication
Any security caveats for Playwright MCP?
Browser actions run in the connected session. Use an isolated profile for testing and do not attach personal or production-admin sessions.
Where is the Playwright MCP source?
Tools highlighted here include browser_navigate, browser_snapshot, browser_click, browser_type (21 total). Source and updates: https://github.com/microsoft/playwright-mcp
Playwright MCP: Setup, Auth & 21 Tools | Bowora