RAG Web Browser Actor MCP

How to Install and Use RAG Web Browser Actor MCP

If you want to use the RAG Web Browser Actor MCP, this guide will help you install and use it in a few simple steps. The RAG Web Browser Actor MCP lets you perform web searches or fetch webpage content quickly. Let's get started!

What You Need Before Installation

Before you begin, make sure you have some important things ready. You need to have Node.js (version 18 or higher) installed on your computer. You also need an Apify API Token, which is like a secret key that allows access to Apify services.

With those ready, you can now install the RAG Web Browser Actor MCP.

Step 1: Clone the MCP Server Repository

First, you need to download the MCP server code to your computer. You do this by cloning the GitHub repository. Open your command prompt or terminal and type:

git clone https://github.com/apify/mcp-server-rag-web-browser.git
cd mcp-server-rag-web-browser
npm install

This command will copy the complete MCP server files and install all the necessary packages.

Step 2: Build the MCP Server

Next, you need to build the server to prepare it for running. Use the following commands:

npm run build

This will compile the source files and make the MCP ready to start.

Step 3: Set Your Apify Token

The MCP server needs your Apify API Token to work. You must set it as an environment variable by running:

export APIFY_TOKEN=your-apify-api-token

Replace your-apify-api-token with your actual token.

Step 4: Run the MCP Server

After setting the token, you can run the server. Use this command:

npx @apify/mcp-server-rag-web-browser

The server will start and listen for commands. This MCP server works locally and handles requests by querying Google search or fetching webpages directly.

How to Use RAG Web Browser Actor MCP Commands

The main tool offered by this MCP is called search. It takes a search query or URL and returns cleaned content.

Example command for searching Google with a query:

{
  "tool": "search",
  "requests": [
    {
      "query": "latest technology news",
      "maxResults": 3,
      "scrapingTool": "raw-http",
      "outputFormats": ["markdown"]
    }
  ]
}

• query is your search term or a direct URL.
• maxResults controls how many search results to fetch.
• scrapingTool can be raw-http (fast for static pages) or browser-playwright (better for sites with a lot of JavaScript).
• outputFormats tells the server to format the response in Markdown.

Example for fetching a single URL:

{
  "tool": "search",
  "requests": [
    {
      "query": "https://example.com",
      "outputFormats": ["markdown"]
    }
  ]
}

Here the MCP fetches the content from the exact website you give.

Tips and Alternatives

• The RAG Web Browser Actor MCP is now deprecated. Instead, you can use the recommended online service at mcp.apify.com for easier setup and more features.
• Using mcp.apify.com means no local setup and instant access to thousands of Apify Actors including this one.
• For more control, the MCP server can be integrated directly with your AI tools using its standard input/output streams.

That’s all! Now you can install and operate the RAG Web Browser Actor MCP to search the web or fetch webpage content efficiently. Just follow these steps and start exploring information online with your AI tools.