Concurrent Browser

A concurrent browser MCP server that supports multiple parallel browser instances, based on Playwright.
  • typescript

8

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "sailaoda-concurrent-browser-mcp": {
      "command": "npx",
      "args": [
        "concurrent-browser-mcp",
        "--max-instances",
        "20"
      ]
    }
  }
}

You run a concurrent browser MCP server to automate and manage multiple browser instances in parallel. This server gives you reliable instance management, flexible configuration, and robust resource cleanup, enabling scalable browser automation across Chromium, Firefox, and WebKit.

How to use

You use an MCP client to control the server by issuing commands that create, manage, and interact with browser instances. Start by choosing how you want to run the server locally (via npx, npm install, or a local build). Then configure the client to point at your running MCP server, selecting a maximum number of concurrent instances and any browser preferences you need. You can create multiple browser instances, navigate pages, interact with elements, take screenshots, and retrieve page information. The server automatically handles instance lifecycle, including cleanup of timed-out instances, so you don’t have to manage resources manually.

How to install

Prerequisites: ensure you have Node.js installed on your system. You should also have a working npm or npx setup to run PMC commands.

Install from npm (Recommended)

# Global installation
npm install -g concurrent-browser-mcp

# Or run directly without installation
npx concurrent-browser-mcp

Build from source

# Clone repository
git clone https://github.com/sailaoda/concurrent-browser-mcp.git
cd concurrent-browser-mcp

# Install dependencies
npm install

# Build project
npm run build

# Optional: Global link (for local development)
npm link

Quick start with a quick script

You can start the server using a quick start script once you have the project available in your environment.

MCP client configuration (server references)

Configure your MCP client to launch the server using the commands shown below. Each configuration option is a self-contained MCP entry that you can reference from your client configuration.

General startup patterns

  • Start with the default configuration: use the command to run the MCP server with standard options. - Customize by selecting a maximum number of instances, a preferred browser, and headless mode.

Additional configuration notes

The server supports multiple ways to run the MCP server locally. You can use npx for quick execution, install the package globally for repeated use, or run a local build directly from your source tree. The server options include instance limits, timeouts, browser type, headless mode, viewport size, and proxy settings.

Security and proxy considerations

Proxy configuration is supported to help you operate browser automation behind a proxy. You can specify a proxy on the command line, rely on automatic local proxy detection, or configure a proxy in the MCP configuration. Proxy detection runs at startup and can display the detected proxy in startup logs.

Troubleshooting and notes

If you encounter startup or runtime issues, check that the Node.js version is compatible, ensure dependencies are installed, and verify the command and path to any built artifacts. For local builds, confirm that the dist/index.js path exists and that the working directory is correct when starting the server.

Examples and quick references

Examples show how to create multiple browser instances, navigate pages, take screenshots, and retrieve page information. Use parallel calls to initialize several instances and perform actions concurrently to maximize throughput.

Notes

This MCP server supports multiple browser instances running concurrently, with automatic cleanup of timed-out instances and flexible configuration for different browser types. It provides complete browser automation capabilities such as navigation, clicking, input, and screenshots across Chromium, Firefox, and WebKit.

Available tools

browser_create_instance

Create a new browser instance with specified options such as browser type, headless mode, and viewport.

browser_list_instances

List all currently managed browser instances.

browser_close_instance

Close a specific browser instance by its identifier.

browser_close_all_instances

Close all running browser instances.

browser_navigate

Navigate an instance to a specified URL.

browser_go_back

Navigate back in the browser history.

browser_go_forward

Navigate forward in the browser history.

browser_refresh

Refresh the current page.

browser_click

Click on a page element using a selector.

browser_type

Type text into a field identified by a selector.

browser_fill

Fill a form field with a value.

browser_select_option

Select a dropdown option.

browser_get_page_info

Retrieve detailed information about the current page.

browser_get_element_text

Get text content of a specified element.

browser_get_element_attribute

Get attributes of a specified element.

browser_screenshot

Capture a screenshot of the current page.

browser_get_markdown

Get the current page or content as Markdown.

browser_wait_for_element

Wait for an element to appear in the DOM.

browser_wait_for_navigation

Wait for page navigation to complete.

browser_evaluate

Execute JavaScript in the context of the page.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational