To connect an MCP server to browser tabs you already have open, the most direct documented route is Playwright MCP in extension mode: install the Playwright Extension in Chrome or Edge, add npx @playwright/mcp@latest --extension to your MCP client configuration, then restart or reload the client and approve the browser connection. This can reuse the selected profile’s logged-in sessions, cookies, tabs, and installed extensions—so treat the connection as access to that browser context, not as a harmless read-only link.
Choose the connection method for the browser you want to use
An MCP browser server gives an AI client a way to interact with a browser. The right setup depends on whether you need the tabs and profile you already use, Chrome’s debugging capabilities, or Firefox automation. These options are related, but they are not interchangeable.
| Method | Browser and session | Connection | Good fit |
|---|---|---|---|
| Playwright MCP + Extension | Existing Chrome or Edge profile and tabs | Install the extension; start Playwright MCP with --extension |
Using logged-in tabs, SSO or 2FA flows, and pages that rely on installed extensions |
| Chrome DevTools MCP | Existing or newly launched Chrome | --autoConnect or a browser URL exposed through remote debugging |
DevTools inspection, debugging, and live Chrome sessions |
| Firefox DevTools MCP | Firefox | Firefox automation through WebDriver BiDi or Classic protocols | Firefox-specific automation and debugging |
If your main goal is to give an agent access to the Chrome or Edge profile you are already using, start with Playwright’s extension mode. Chrome DevTools MCP is another route for Chrome, but its automatic connection has specific Chrome requirements. Mozilla’s Firefox DevTools MCP is a separate Firefox-focused server, not a way to install the Playwright extension workflow in Firefox.
Connect Playwright MCP to existing Chrome or Edge tabs
Playwright extension mode connects the MCP server to a browser profile where the Playwright Extension is installed. The extension can let the server work with existing tabs and reuse the profile’s authenticated state and extensions. That is useful when a site requires a login, SSO, 2FA, or a browser extension, but it also means the agent may be operating in a context that contains sensitive information.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11#1 Best Overall
1. Install the extension in the intended browser profile
Install the Playwright Extension in Chrome or Edge, using the profile you want the agent to access. If you keep separate profiles for work, personal browsing, or testing, install it only in the profile appropriate for the task. Installing the extension in one profile does not mean another profile will have the same tabs or login state.
2. Add Playwright MCP to your MCP client configuration
In the MCP client’s server configuration, add this entry. The JSON shape below uses the common mcpServers key; use the configuration file and reload process your particular MCP client expects.
{
"mcpServers": {
"playwright": {
"command": "npx",
"args": ["@playwright/mcp@latest", "--extension"]
}
}
}The command asks npx to run the Playwright MCP package and start it in extension mode. The @latest tag requests the latest published package when invoked; if your environment manages package versions centrally, follow that environment’s normal package policy.
3. Reload the MCP client and connect the browser
Save the configuration, then start or reload the MCP client so it reads the new server entry. Allow the extension to connect to the intended browser profile. If your client or browser displays a permission request, check that it is for the expected connection before approving it. The extension is documented to attach to the last-used browser profile that has the extension installed.
Free tools Windows power users keep installed
One-click scans. No signup required.
4. Select a profile explicitly when necessary
If more than one browser profile has the extension installed, the default last-used-profile behavior may select a profile other than the one you intended. Add the profile directory name to the arguments:
{
"mcpServers": {
"playwright": {
"command": "npx",
"args": [
"@playwright/mcp@latest",
"--extension",
"--profile-dir-name=Profile 1"
]
}
}
}Replace Profile 1 with the actual profile directory name you want to target. This setting selects a profile; it does not create a profile, install the extension there, or transfer the login state from another profile.
Connect Chrome DevTools MCP to Chrome instead
Chrome DevTools MCP connects an AI agent to a live Chrome instance. Chrome’s automatic connection instructions require Chrome 144 or later and enabling remote debugging at chrome://inspect/#remote-debugging. The MCP server entry is:
{
"mcpServers": {
"chrome-devtools": {
"command": "npx",
"args": ["chrome-devtools-mcp@latest", "--autoConnect"]
}
}
}After adding it, reload the MCP client and follow Chrome’s connection prompt. Approve the session only if you trust both the MCP client and the agent that will use it. Chrome documents that an automatic connection can expose open tabs and data accessible through JavaScript APIs, including session storage, local storage, and cookies.
Manual remote-debugging connection
If you prefer to start a browser with a debugging endpoint yourself, launch Chrome with a debugging port and a separate user-data directory, then tell the MCP server where to connect:
google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-profile-stable{
"mcpServers": {
"chrome-devtools": {
"command": "npx",
"args": [
"chrome-devtools-mcp@latest",
"--browser-url=http://127.0.0.1:9222"
]
}
}
}The launch command uses a separate directory rather than your everyday Chrome profile. The MCP server’s --browser-url value must point to the debugging endpoint you started. Chrome also documents a WebSocket endpoint option, --ws-endpoint, for clients that use a browser WebSocket URL; use that only when your client and setup are configured for that endpoint.
Use Firefox’s MCP server for Firefox automation
Mozilla documents firefox-devtools-mcp as an MCP server for automating and interacting with Firefox through WebDriver BiDi and Classic protocols. It is a Firefox automation-server path. The documented workflow is not the Playwright Extension setup for existing Chrome or Edge tabs, so choose it when the target browser is Firefox rather than expecting Chrome/Edge extension behavior to carry over.
Keep the browser session and MCP server secure
A browser connection can carry more authority than the MCP tool list alone suggests. An agent working through an attached profile may be able to interact with authenticated pages and browser data visible in that context. Chrome specifically warns that auto-connect may expose tabs, cookies, session storage, local storage, and other information surfaced through JavaScript APIs.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitches- Use a trusted client and agent. Do not approve a connection just because a prompt appears; confirm which client and browser session are requesting access.
- Separate sensitive work. Prefer a dedicated browser profile for agent tasks when possible. Avoid attaching a profile that contains unrelated personal, financial, or work accounts.
- Choose the profile deliberately. With Playwright extension mode, check which profile has the extension and whether the last-used-profile behavior will target the right one. Use
--profile-dir-namewhen you need to specify it. - Treat Playwright MCP configuration as privileged tooling. Playwright’s MCP documentation warns that the server can run arbitrary JavaScript in the Playwright server process and is equivalent to remote-code-execution capability for a trusted client. Do not add or run an untrusted server entry.
Profile isolation reduces accidental exposure, but it is not a substitute for trusting the software and agent you connect. The practical boundary is the browser context and server process you have chosen to make available.
Understand what this setup does—and what WebMCP does not do
An extension or debugging bridge connects an MCP server to a browser instance or profile. WebMCP is a different browser-focused technology: Chrome’s comparison says it is not an extension or a replacement for MCP. It should not be used as the name for this connection workflow. If you want an agent to work with existing tabs, use one of the browser-server paths above; do not expect WebMCP itself to attach the agent to your browser profile.
Troubleshoot common connection problems
- The MCP server does not appear in the client. Check that the JSON is valid, that the entry is inside the client’s expected
mcpServersobject, and that you restarted or reloaded the client after saving. MCP clients can differ in where they store configuration. - Playwright MCP starts, but the wrong browser profile is used. The extension attaches to the last-used profile that has it installed. Install the extension in the intended profile and, if multiple profiles have it, set
--profile-dir-name=Profile 1to the correct directory name. - The agent cannot see the tabs you expected. Confirm the tabs are open in the selected profile, not another profile or a different browser. The extension workflow targets Chrome or Edge profiles where the extension is installed.
- Chrome automatic connection is unavailable. Confirm that Chrome is version 144 or later and that remote debugging has been enabled at
chrome://inspect/#remote-debugging. If automatic connection is not suitable, use the documented manual remote-debugging approach. - The manual Chrome connection fails. Make sure Chrome was launched with the remote-debugging port and user-data directory, and that
--browser-urlpoints to the matching local endpoint. A server pointed at a different port or endpoint cannot attach to the launched browser. - A login or page still does not work as expected. Verify that the selected profile is actually signed in and that the needed browser extension is installed and enabled in that same profile. Extension mode can reuse the profile’s state; it does not make one profile inherit another profile’s cookies or installed extensions.
- You are using Firefox but following Chrome or Playwright extension steps. Switch to Mozilla’s Firefox DevTools MCP path for Firefox automation; it uses WebDriver protocols rather than the described Chrome/Edge extension connection.
Or skip the browser setup
If your goal is to capture a website image or PDF rather than let an agent operate your open browser tabs, ScreenshotNeo is a separate screenshot API and MCP server—not a bridge to your existing browser extension. A single GET request can return a PNG, JPEG, WebP, or PDF. For example, this cURL request captures a page as WebP:
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webpSee the ScreenshotNeo API documentation for setup and options. The service accepts consent banners as a visitor and removes 60+ known consent platforms, newsletter popups, and chat widgets before capture; each of those steps can be turned off. Bot checks or CAPTCHAs, blank pages, timeouts, failed loads, and cache hits are not billed, and each response identifies the page verdict and billing status in headers. Its MCP server offers take_screenshot, get_page_info, and capture_pdf for Claude, Cursor, and other MCP clients.
ScreenshotNeo is for screenshot and PDF capture, not controlling authenticated tabs through your installed browser profile. It includes full-page captures with lazy images loaded, CSS-selector element capture, dark mode, device presets and custom viewports, retina scale, PDF settings, HTML/CSS rendering, custom CSS and JavaScript, clicks, selector hiding, wait conditions, request blocking, custom headers and cookies, user-agent and Authorization settings, timezone and geolocation, transparent backgrounds, resizing, configurable cache TTL, signed image links, asynchronous jobs with signed webhooks, bulk captures of up to 100 URLs per call, a usage API, and an OpenAPI spec. Common screenshot API parameter names also work to make switching easier.
| Plan | Monthly captures | Price |
|---|---|---|
| Free | 1,000 | $0; no card |
| Starter | 3,000 | $5 |
| Growth | 15,000 | $15 |
| Pro | 60,000 | $39 |
| Scale | 250,000 | $99 |
| Business | 1,000,000 | $249 |
Yearly billing gives two months free, and every feature is available on every plan. Sign up for ScreenshotNeo to get 1,000 screenshots a month free with no card.
Frequently Asked Questions
Does Playwright extension mode work with both Chrome and Edge?
The documented extension workflow is for Chrome or Edge profiles where the Playwright Extension is installed.
Is WebMCP the same thing as connecting an MCP server to my browser?
No. WebMCP is a browser-focused technology for websites to expose capabilities to agents; it is not the extension or debugging bridge described here.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Can I use ScreenshotNeo to give an agent control of my open browser tabs?
No. ScreenshotNeo is for website screenshots and PDFs; it does not attach to an existing browser profile or control its tabs.
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.
