@onpiste/mcp

A Model Context Protocol (MCP) server for browser automation that enables AI applications to control and interact with web browsers through a Chrome extension.

✨ Features

Core Capabilities

• 🤖 AI-Powered Task Execution: Let AI agents handle complex multi-step browser tasks autonomously
• 🌐 Advanced Navigation: Full control over browser navigation including URL navigation, back/forward, and page refresh
• 🖱️ Smart Interactions: Click elements, type text, select dropdown options, and press keyboard keys
• 📜 Flexible Scrolling: Scroll by pixels, percentage, or directly to specific text on the page
• 🗂️ Tab Management: Open, close, switch between tabs, and get information about all open tabs
• 📸 Visual Capture: Take screenshots and capture accessibility snapshots of web pages
• 🔍 State Inspection: Access page URLs, titles, console logs, and full browser state information
• ⏱️ Smart Waiting: Wait for specific time durations or for elements to appear on the page

Key Advantages

• ⚡ Fast: Automation happens locally on your machine, resulting in better performance without network latency
• 🔒 Private: Since automation happens locally, your browser activity stays on your device and isn't sent to remote servers
• 👤 Logged In: Uses your existing browser profile, keeping you logged into all your services
• 🥷🏼 Stealth: Avoids basic bot detection and CAPTCHAs by using your real browser fingerprint
• 🤖 AI-Ready: Compatible with AI applications like Cursor, Claude Code, VS Code, Windsurf, Zed, and more

🛠️ Installation

Requirements

• Node.js >= v18.0.0
• Cursor, Claude Code, VS Code, Windsurf, Zed or another MCP Client
• Chrome or Chromium-based browser with the @onpiste extension installed

Quick Start

1. Install the @onpiste Chrome extension (if not already installed)
2. Configure your AI application using one of the methods below

{
  "mcpServers": {
    "onpiste": {
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

claude mcp add onpiste -- npx -y @onpiste/mcp@latest

{
  "mcpServers": {
    "onpiste": {
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

"mcp": {
  "servers": {
    "onpiste": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

{
  "mcpServers": {
    "onpiste": {
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

{
  "context_servers": {
    "onpiste": {
      "source": "custom",
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

{
  "mcpServers": {
    "onpiste": {
      "command": "npx",
      "args": ["-y", "@onpiste/mcp@latest"]
    }
  }
}

🔨 Available Tools

The MCP server provides a comprehensive set of browser automation tools:

🤖 AI Task Execution

• run_task - Request the agent to complete a task within the browser environment
• navigate_to - Navigate to a specified URL with onpiste optimizations

🧭 Navigation

• browser_go_back - Go back to the previous page
• browser_go_forward - Go forward to the next page
• browser_refresh - Refresh the current page

🖱️ Interaction

• browser_click - Perform click on a web page element
• browser_type - Type text into an editable element
• browser_select_option - Select an option in a dropdown

⌨️ Keyboard Operations

• browser_press_key - Press a single key on the keyboard
• browser_send_keys - Send keys to the page (supports key combinations like Control+A)

📜 Scrolling

• browser_scroll_to_percent - Scroll to a percentage of the page or element
• browser_scroll_by - Scroll by a specified number of pixels
• browser_scroll_to_text - Scroll to specific text on the page

🗂️ Tab Management

• browser_get_tabs - Get information about all open browser tabs
• browser_switch_tab - Switch to a different browser tab
• browser_open_tab - Open a new browser tab
• browser_close_tab - Close a browser tab

🔍 State Inspection

• browser_snapshot - Capture accessibility snapshot of the current page
• browser_get_url - Get the current page URL
• browser_get_title - Get the current page title
• browser_get_console_logs - Get the console logs from the browser
• browser_get_state - Get the current browser state including page information

📸 Screenshots

• browser_screenshot - Take a screenshot of the current page

⏱️ Wait Operations

• browser_wait - Wait for a specified time in seconds
• browser_wait_for_element - Wait for an element to appear on the page

📝 Usage Examples

Once configured, you can interact with your browser through natural language commands in your AI assistant:

Navigate to github.com and search for "MCP servers"

Take a screenshot of the current page

Click on the login button and fill in the form

The MCP server will automatically handle the browser automation through the @onpiste extension.