Yandex Browser Tabs

You run a specialized MCP server to control Yandex Browser tabs with extended capabilities. It lets you access the browser console, scroll to elements, extract text and HTML, read attributes and meta information, and perform interactive actions like hovering, clicking, typing, and waiting for elements. This guide walks you through using and installing the server so you can automate browser tasks reliably.

How to use

To use the MCP server, you run a local MCP client that connects to the server process. The server exposes a set of practical actions you can invoke from your client to control Yandex Browser tabs, query page data, and interact with page elements. You will connect to an already-open browser instance or start the browser with a debugging port, then issue commands to navigate, extract content, or simulate user interactions.

How to install

Prerequisites you need before installing: Node.js and npm installed on your computer. You will clone the project, install dependencies, and build the project before use.

# Clone the project
git clone <repository-url>
cd yandex-browser-mcp

# Install dependencies
npm install

# Build the project
npm run build

Additional sections

Configuration, usage notes, and examples help you get the most out of the MCP server. Below you’ll find the recommended workflow for running the browser in debug mode, and how to connect the MCP server to your client configuration.

Run the browser in debug mode using the provided launcher or manually start the browser with a remote debugging port.

Example client configuration to connect to the MCP server:

{
  "mcpServers": {
    "yandex-browser": {
      "command": "node",
      "args": ["C:\\Users\\Professional\\Desktop\\yandex-browser-mcp\\build\\index.js"]
    }
  }
}

Notes on capabilities

The MCP server supports a comprehensive set of actions for browser control and data extraction, including connecting to a running browser, listing tabs, navigation, element interaction (click, type, hover), scrolling, getting text and HTML, reading attributes, obtaining page info, and taking screenshots. It also includes utilities for waiting for elements, filling forms, and evaluating scripts.

Available tools

connect_to_browser

Connect to an already opened Yandex Browser instance, establishing a control session for subsequent actions.

list_tabs

Retrieve a list of all currently open browser tabs to identify targets for actions.

navigate

Navigate a specified tab to a given URL, optionally waiting for a selector to appear.

click

Perform a click action on an element identified by a CSS selector within a tab.

type

Type text into an input field selected by a CSS selector, with optional delay.

hover

Move the mouse over a target element to trigger hover effects.

key_press

Simulate a keyboard press or a key combination within a tab.

scroll

Scroll the page or a specific element in a chosen direction with optional smooth scrolling.

get_text

Extract text content from the page or a selected element.

get_html

Retrieve HTML content, with options for inner or outer HTML.

get_attributes

Read attributes from a target element, filtered by a list of attribute names.

get_page_info

Obtain page-level metadata such as URL, title, description, and meta tags.

get_console_logs

Fetch console log entries by type (log, warn, error, info) with an optional limit.

fill_form

Populate forms by providing field selectors and corresponding values.

wait_for_element

Await the appearance of a target element with optional timeout and visibility constraints.

screenshot

Capture a screenshot of the page or a specific element, with options for full page capture.

evaluate

Execute arbitrary JavaScript within the context of the page.

execute_test

Run a predefined test by selecting answers and submitting through a specified button.