MCP Live Data Acquisition Module

A robust implementation of the Model Context Protocol (MCP) designed to provide instant access to current external datasets through interchangeable search backends. It is presently configured to leverage the Brave Search API.

✨ Core Capabilities

• Up-to-the-Minute Fact Retrieval: Equips automated assistants with the ability to pull the freshest web data.
• Interchangeable Sourcing: Modular design facilitates effortless substitution of underlying search engines.
• Uniform Data Schema: Guarantees search outcomes are delivered in a clean, predictable JSON structure.
• Intelligent Query Routing: Automatically classifies incoming informational requests to offer contextually precise algorithmic direction.

📋 Prerequisites

• Runtime Environment: Node.js version 16.x or above
• Package Manager: npm version 7.x or higher
• API Credential: A valid Brave Search API Key is mandatory for service operation

🚀 Setup Procedure

1. Source Code Retrieval: bash git clone https://github.com/gabrimatic/mcp-web-search-tool.git cd mcp-web-search-tool

2. Dependency Installation: bash npm install

3. Configuration Setup: Establish a .env file in the root directory:

BRAVE_API_KEY=your_secret_key MAX_RESULTS=15 # Override default result limit (10) REQUEST_TIMEOUT=15000 # Customization for request termination delay (Default 10s)

1. Build Artifact Generation: bash npm run build

💻 Operational Use

Server Initialization

bash npm start

Verification Routine

bash node test-server.js

Integration with the Claude Desktop Client (Optional Enhancement)

A significant benefit of this module is its native compatibility with the Claude Desktop application, enabling real-time data lookups for queries that demand current context.

Configuration Steps

1. Produce a configuration file named claude_desktop_config.json:

{ "mcpServers": { "mcp-web-search": { "command": "node", "args": [ "/absolute/path/to/your/mcp-web-search-tool/build/index.js" ] } } }

1. Launch the Claude Desktop application utilizing this specific configuration file.

2. Pose questions to Claude that inherently require current external knowledge; the system will automatically invoke this search utility.

🎥 Demonstration Video

Witness the live synergy between Claude and the MCP Web Search Tool for immediate, AI-driven web navigation:

📺 Claude + MCP Web Search – Live Demo

Illustrative Inquiry Examples:

• "What is the current consensus among analysts regarding the leading contenders in the NBA MVP selection following recent fixtures?"
• "Provide the most recent developments in the field of advanced artificial intelligence research."
• "What is the prevailing meteorological condition in Metropolitan New York City at this moment?"
• "Report on the immediate performance status of key global equity indices."

🛠️ Exposed Functionality

Data Retrieval Engine (web_search)

• Function Identifier: web_search
• Purpose: To query the live internet for CURRENT facts and data.
• Mandatory For: Checking weather, trending topics, athletic results, market valuations.
• Input Arguments:
• search_term (Text): The precise string to execute the lookup against.
• provider (Text, Optional): Specifies the backend engine to use (Brave is default).

Automated Classification Buckets

This utility autonomously segments input into the following knowledge domains: - Meteorological forecasts - Breaking news and current affairs - Sports standings and scores - Financial market activity and data - Time-sensitive factual queries - Broad-scope informational seeking

📜 Licensing Details

MIT License

👨‍💻 Creator

Authored by Hossein Yousefpour

© All intellectual property rights reserved.

📝 Extended Analysis (Medium)

For an in-depth exploration of this module's architecture, capabilities, and how it augments AI data acquisition strategies, consult our dedicated Medium publication: 📖 Deep Dive into MCP Web Search Tool

☕ Patronage