Traylinx Search Engine MCP Server

🚀 Traylinx Search Engine MCP Server

A Model Context Protocol (MCP) server that serves as a bridge for the deployed Agentic Search API. It enables MCP clients like Claude Desktop and Cursor to leverage intelligent search capabilities, providing detailed text summaries and structured data (HTML, images, etc.).

🚀 Quick Start

Traylinx Search Engine MCP Server is a bridge connecting MCP clients with the Agentic Search API. The search tool allows MCP clients to perform web searches and obtain detailed results. The workflow involves receiving search queries, invoking the Agentic Search API, and processing and formatting responses for MCP clients.

✨ Features

• It serves as a bridge between MCP clients and the Agentic Search API.
• Allows MCP clients to utilize intelligent search functions and provides detailed text summaries and structured data.
• Supports multiple content types and time filters for search results.

📦 Installation

Environment Variable Configuration

Set the following environment variables to configure the server behavior:

• AGENTIC_SEARCH_API_KEY: The API key for the Agentic Search API, required.
• AGENTIC_SEARCH_ENDPOINT: Customize the endpoint URL of the Agentic Search API. The default value is the default endpoint.

Example Configuration File

{
  "environmentVariables": {
    "AGENTIC_SEARCH_API_KEY": "your_api_key_here",
    "AGENTIC_SEARCH_ENDPOINT": "https://api.agents.com/v1"
  }
}

💻 Usage Examples

Basic Usage

search

Use Traylinx's API to perform web searches. This API provides detailed and context - relevant results with citations. By default, no time filter is applied. Input:

• query (string): The search query to be executed.
• search_recency_filter (string, optional): Filter search results by time. Options: "month", "week", "day", "hour". If not specified, no time filter is applied by default.

Advanced Usage

Multi - Content Type Return

The server supports the following content types:

• Text Summaries: Concise search results.
• Structured Data: URLs, titles, and descriptions.
• News Articles: Articles with timestamps and thumbnails.

Time Filter Usage

{
  "name": "search",
  "arguments": {
    "query": "SpaceX launches",
    "search_recency_filter": "week"
  }
}

📚 Documentation

How It Works

1. Project Introduction: The Traylinx Search Engine MCP Server acts as a bridge between MCP clients and the Agentic Search API.
2. Tool Explanation: The search tool allows web searches to be performed from MCP clients and returns detailed results.
3. Workflow:
   • Receive search queries.
   • Invoke the Agentic Search API.
   • Process and format responses for MCP clients.

Verification Methods

Manual Testing

Run the following command to verify that the server is working properly:

curl -X POST http://localhost:3000/api/search \
     -H "Content-Type: application/json" \
     -d '{"query": "latest tech news", "search_recency_filter": "day"}'

Automated Test Script

const fetch = require('node-fetch');

async function testSearch() {
    const response = await fetch('http://localhost:3000/api/search', {
        method: 'POST',
        headers: {
            'Content-Type': 'application/json'
        },
        body: JSON.stringify({
            query: "latest tech news",
            search_recency_filter: "day"
        })
    });
    
    const data = await response.json();
    console.log(data);
}

testSearch();

Deployment Information

Deployment on Smithery.ai

Deploy on Smithery.ai:

1. Log in/Register for a Smithery account.
2. Click "New MCP Server".
3. Enter the ID: traylinx-search-engine-mcp-server.
4. Set the base directory to the root directory.
5. Click "Create".

Troubleshooting

• Problem 1: Unable to connect to the API?
  • Check the network connection and the API address for correctness.
• Problem 2: No search results?
  • Confirm that the API key is valid and correctly configured.
• Logging:
  • Set LOG_LEVEL to DEBUG to get detailed log information.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.