Free tools Windows power users keep installed
One-click scans. No signup required.
To let Claude Code open a browser and interact with a website, register Playwright MCP with one terminal command, then restart Claude Code. The server can navigate to a page, return an accessibility snapshot, and let the agent act on identified elements before checking the updated page. You need Node.js 20 or newer for the current Playwright MCP installation instructions.
Set up Playwright MCP in Claude Code
- Check your Node.js version. Playwright’s current MCP installation guide requires Node.js 20 or newer. The browser downloads automatically the first time the server is used. Playwright MCP installation guide.
- Register the server. In a terminal where the Claude Code command is available, run:
claude mcp add playwright npx @playwright/mcp@latestThis is the command documented in the Playwright MCP README and Claude Code client instructions.
- Start a new Claude Code session. The server becomes available in the next session, so restart Claude Code if it was already open.
- Try a harmless demo flow. Ask Claude Code to navigate to
https://demo.playwright.dev/todomvc, add a todo item such as “Check Playwright MCP,” and report what changed. This follows the example flow in the installation guide.
How Claude Code inspects and clicks through a page
Playwright MCP is a server that exposes browser navigation and interaction tools through MCP. Its typical loop is inspect, act, then inspect again: the agent navigates to the page, receives a structured accessibility snapshot, uses references in that snapshot to identify controls, and then clicks, types, or fills a field. It can request another snapshot to see the resulting state. The Playwright MCP introduction describes this interaction model and the available tools.
As an Amazon Associate I earn from qualifying purchases.
- Navigate: Open the TodoMVC demo URL.
- Inspect: Read the accessibility snapshot and identify the todo textbox from its role and label.
- Act: Use the textbox reference to enter an item and submit it, for example by pressing Enter.
- Verify: Inspect the updated page state and report whether the new todo appears.
This approach does not require the agent to rely on screenshots for the basic interaction. A page’s structure, authentication requirements, or behavior can still affect whether a particular flow works as expected.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesChoose whether the browser is visible and where it runs
By default, Playwright MCP launches a headed browser, so you can watch the session. To run without a visible browser window, add --headless to the server arguments. The configuration options also support selecting Chrome, Firefox, WebKit, or Microsoft Edge. See Playwright MCP configuration options for the exact arguments.
#1 Best Overall
The basic command uses a local browser. For an advanced setup, Playwright MCP can connect to a remote browser using a Playwright endpoint or a CDP endpoint; this is not needed for the local demo flow. Endpoint details and options are in the configuration documentation.
Decide how browser state should be saved
Playwright MCP uses a persistent browser profile by default, which preserves cookies and login state between sessions. Choose this behavior deliberately, especially on a shared machine or when working with sensitive accounts.
- Persistent profile: Useful when a workflow needs a login to remain available between sessions.
- Isolated mode: Starts a fresh session and does not save browser state to disk. You can provide storage state when an isolated session needs preloaded authentication.
These profile and storage-state options are documented in the Playwright MCP configuration guide. Treat saved authentication state as sensitive: avoid sharing it, and use the mode that matches the task.
When to choose MCP instead of Playwright CLI
Playwright positions MCP for specialized agentic loops that benefit from persistent state and iterative reasoning about page structure. It positions Playwright CLI as a more token-efficient command workflow for coding agents working with large codebases. This is the project’s qualitative guidance, not a quantified performance comparison. See the Playwright MCP introduction.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
