mcp-server-rag-web-browser
About This Server
A MCP Server for the RAG Web Browser Actor. This Actor serves as a web browser for large language models (LLMs) and RAG pipelines, similar to a web search in ChatGPT.
Server Information
This webpage provides information about the Model Context Protocol (MCP) Server for the RAG Web Browser Actor, implemented by Apify. The MCP Server allows AI applications, such as Claude Desktop, to connect to external tools and data sources, specifically enabling web search and content scraping functionalities for large language models (LLMs) and RAG pipelines. The page includes setup and configuration instructions, development details, and debugging guidance.
ā Key Points:
- The project implements an MCP server for the RAG Web Browser Actor.
- The Actor allows AI to perform web searches and scrape content from URLs.
- The server offers tools to query Google Search and fetch cleaned content as Markdown.
- Configuration prerequisites include MacOS or Windows, Claude Desktop, Node.js, and an Apify API token.
- The server supports testing via a simple local client (stdio) and a chat local client (stdio).
- The server is testable using the Server-Sent Events (SSE) transport.
- MCP Inspector can be used for debugging purposes.
- The MCP server enables AI agents to interact with the web through the RAG Web Browser Actor.
- It facilitates web searches and extracts content from specified URLs.
- The setup requires specific software and API keys.
- Testing and debugging instructions are provided using local clients and MCP Inspector.
- Tools:
- `search`: Queries Google Search, scrapes the top N URLs, and returns cleaned content as Markdown, including arguments for the search query and maximum results.
- Setup:
- Requires cloning the repository, installing dependencies, and building the project.
- Claude Desktop configuration involves modifying the `claudedesktopconfig.json` file to include the MCP server details.
- SSE transport testing instructions are provided.
- Debugging:
- The instructions guide the debugging process using the MCP Inspector, including launching the Inspector and building the mcp-server-rag-web-browser package.
- Repository Information:
- Owner: apify
- Name: mcp-server-rag-web-browser
- Public
- 59 stars, 5 forks
- Main language: TypeScript
- License: Apache-2.0
š Main Findings:
š Details:
šÆ Conclusion:
The webpage details a functional MCP server project designed to extend the capabilities of AI models by providing web browsing functionalities. It offers a comprehensive guide for setting up, testing, and debugging the server, enabling developers to integrate web search and content extraction into their AI applications.
Server Features
search
Query Google Search, scrape the top N URLs from the results, and returns their cleaned content as Markdown.
Provider Information
Apify
cloud Provider