| name | playwright-cli |
| description | Browser automation via @playwright/mcp (Microsoft). Use this when the user wants to navigate websites, fill forms, take screenshots, scrape web content, test web apps, or run any multi-step browser workflow. Requires no display (headless mode supported).
|
| license | MIT |
| compatibility | Requires: npx (@playwright/mcp) |
playwright-cli
Browser automation via @playwright/mcp (Microsoft). Navigate websites,
fill forms, take screenshots, and run complex web automation workflows.
Setup
@playwright/mcp is available via npx. No global install required.
When to use this skill
- Automate website interactions (clicking, scrolling, forms)
- Take screenshots or generate PDFs of web pages
- Web scraping when no API is available
- Test web applications
- Multi-step workflows on websites
Execution
Start the MCP server:
npx @playwright/mcp@latest --headless
Key flags
--headless — No visible browser window (recommended on servers)--browser chromium — Select browser: chromium / firefox / webkit--allowed-origins https://example.com — Restrict access to specific hosts--port 3000 — Port for the MCP server
Browser agent capabilities
- Navigate: Open URL, back/forward, reload
- Interact: Click, double-click, hover, drag & drop, type, checkboxes, dropdowns
- Read: Text, elements, screenshots (PNG), PDFs
- Tabs: Open, switch, close
- Sessions: Save storage state (cookies, localStorage) for repeated logins
- Request mocking: Intercept and manipulate HTTP requests
Security notes
- Use
--allowed-origins to restrict access where possible
- Use
--headless on servers — no display required
- Storage state files contain cookies/tokens — store securely
Example invocations
Headless browser session (chromium):
npx @playwright/mcp@latest --headless --browser chromium
Restricted origin:
npx @playwright/mcp@latest --headless --allowed-origins https://example.com
With visible browser (local dev):
npx @playwright/mcp@latest --browser chromium
