DICloak Browser MCP is a powerful browser automation and multi-account management tool built on top of the DICloak Anti-Detect Browser. It exposes a rich set of browser interaction capabilities via the Model Context Protocol (MCP), enabling programmatic control over tabs, navigation, content extraction, network inspection, and more—all while maintaining session isolation and digital identity safety.
This MCP server is ideal for developers building AI agents, automation workflows, or scraping tools that require robust, fingerprint-safe browser environments.
✨ Features
- Full browser control: Open tabs, navigate, click, fill forms, capture screenshots, and simulate keyboard input.
- Content extraction: Retrieve HTML, text, element attributes, or semantic embeddings from pages.
- Network monitoring: Capture and analyze network requests with full browser context.
- Interactive element discovery: Find buttons, links, or inputs by selector or fuzzy text match.
- Browser history & console access: Inspect browsing history and JavaScript console output.
- Script injection & communication: Inject custom JS and send messages to injected scripts.
- Semantic search: Index and search page content using vector embeddings.
- Multi-profile & anti-detect support: Leverages DICloak’s core isolation and fingerprint spoofing technology.
💡 All operations are performed within isolated, anti-detection browser profiles—essential for managing multiple accounts safely on platforms like social media, e-commerce, and affiliate networks.
🔌 Integration
To use this MCP server in your application (e.g., with an AI agent framework), add the following configuration to your MCP client:
{
"mcpServers": {
"dicloak-mcp-http": {
"type": "streamableHttp",
"url": "https://mcp.dicloak.com/stream?apikey=XXXXXXXXXXXXXXXXX",
"description": "DICloak Chrome Browser MCP Service"
}
}
}🔑 Replace XXXXXXXXXXXXXXXXX with your actual API key from DICloak.
🛠️ Available Tools
The MCP server provides the following tools (simplified list; full schema available in code):
| Tool Name | Description |
|---|---|
| get_windows_and_tabs | List all open windows and tabs |
| chrome_navigate | Navigate to a URL or refresh |
| chrome_screenshot | Capture full-page or element screenshots (PNG or Base64) |
| chrome_close_tabs | Close tabs by ID, URL, or all |
| chrome_go_back_or_forward | Navigate browser history |
| chrome_get_web_content | Extract HTML or text content (with selector support) |
| chrome_get_element_attributes | Get element properties, styles, and bounding box |
| chrome_click_element | Click by selector or coordinates |
| chrome_fill_or_select | Fill input fields or select dropdown options |
| chrome_get_interactive_elements | Discover clickable/interactive elements |
| chrome_network_request | Send authenticated HTTP requests from browser context |
| chrome_network_capture_start/stop | Capture outgoing network traffic |
| chrome_keyboard | Simulate key presses (e.g., PageDown, Ctrl+C) |
| chrome_history | Search browsing history |
| search_tabs_content | Search open tabs by content |
| chrome_vector_search | Semantic search via embeddings (index, search, etc.) |
| chrome_console | Retrieve console logs and errors |
| chrome_inject_script | Inject or manage JavaScript in the page |
| chrome_send_command_to_inject_script | Communicate with injected scripts via custom events |
| chrome_content_analysis | Summarize, analyze, or extract entities from text |
📝 All tools support optional parameters for fine-grained control. See source code or schema definitions for full specifications.