Aluvia MCP Server: Unblockable Browser Automation for AI Agents

What is Aluvia MCP Server?

The Aluvia MCP Server is a bridge that connects your AI agents (like Claude, Cursor, or Windsurf) to the real web. It implements the Model Context Protocol (MCP) to give Large Language Models (LLMs) the ability to launch browser sessions that are virtually indistinguishable from real human users.

Unlike standard web fetching tools that often get blocked by 403 Forbidden errors, CAPTCHAs, or WAFs (Web Application Firewalls), Aluvia routes traffic through premium US mobile carrier IPs. This allows your agent to access protected sites, gather data, and perform automation tasks without getting blocked.

Why do I need this? (The Problem)

AI agents are powerful, but they struggle to browse the web reliably:

1. Blocking: Agents often use data center IPs, which are flagged immediately by sites like Cloudflare, Akamai, and Datadome.
2. CAPTCHAs: Agents get stuck on "Verify you are human" screens.
3. Geo-Restrictions: Many sites show different content or block access based on location.

Aluvia solves this by routing your agent's requests through real mobile devices (4G/5G/LTE), making the traffic appear organic and trusted.

Key Features

• 🛡️ Auto-Unblocking: The server automatically detects when a request is blocked (403s, challenges) and re-routes it through a clean mobile IP to bypass the block.
• 📱 Mobile Carrier IPs: Access millions of premium US mobile IPs (Verizon, T-Mobile, AT&T) that websites trust.
• ⚡ Smart Routing: To save costs and reduce latency, Aluvia only proxies traffic for domains that actually block you. Open sites are accessed directly.
• 🌍 Geo-Targeting: Instruct your agent to appear from specific regions (e.g., us_ca for California, us_ny for New York) to verify local content or pricing.
• 🔄 IP Rotation: Instantly rotate your IP address with a single tool call if a session gets flagged.
• 🕵️ Block Detection: Advanced heuristics analyze page content, HTTP headers, and WAF signatures to "know" when the agent is being stopped.

Use Cases

1. Market Research & Pricing Intelligence

Scenario: Your agent needs to check competitor pricing on an e-commerce site that blocks scrapers. Solution: Aluvia allows the agent to load the product page as a real mobile user, bypass the anti-bot protection, and extract the correct price.

2. Verified Data Collection

Scenario: You need to verify how ads or search results appear to users in Los Angeles vs. New York. Solution: Use session_set_geo to switch the agent's location and capture screenshots or text from different geo-perspectives.

3. autonomous Web Navigation

Scenario: An agent is navigating a complex travel booking flow but gets stopped by a CAPTCHA halfway through. Solution: Aluvia detects the challenge and rotates the IP or solves the access token, allowing the agent to continue the session without human intervention.

4. Financial & Real Estate Data

Scenario: Aggregating data from strict financial institutions or real estate listings (Zillow, Redfin) that have high security. Solution: Mobile IPs are highly trusted by these platforms, granting access where data center proxies fail.

How to use Aluvia MCP

Prerequisites

1. Node.js (v18 or higher)
2. Aluvia API Key (Get one at dashboard.aluvia.io)

1. Installation

Install the package from npm:

npm install @aluvia/mcp

2. Configuration (Claude Desktop Example)

Add the server to your claude_desktop_config.json:

{
  "mcpServers": {
    "aluvia": {
      "command": "npx",
      "args": ["-y", "aluvia-mcp"],
      "env": {
        "ALUVIA_API_KEY": "your_api_key_here"
      }
    }
  }
}

3. Usage in Chat

Once connected, you can simply ask your agent to perform tasks. The agent will intelligently select the right tools.

Example Prompts:

"Go to example.com and check the latest headlines. If you get blocked, rotate the IP and try again."

"Start a browser session in California (us_ca) and tell me the price of gas on gasbuddy.com."

Available Tools

Your agent will have access to these primitives: