Kagi Search MCP Server

The Kagi Search MCP provides seamless integration with Kagi's search API, allowing AI assistants to perform web searches with higher quality results than traditional search engines. This server enables access to Kagi's privacy-focused search technology, delivering more relevant and less biased search results without tracking or ads. Currently supporting web search functionality, with plans to expand to Kagi's summarization, FastGPT, and news enrichment features in future updates. This MCP enhances AI assistants with up-to-date information from the web through Kagi's superior search technology.

Overview

The Kagi Search MCP server integrates Kagi's search API into AI assistants, providing high-quality, privacy-focused web search capabilities. This allows your AI assistant to retrieve current information from the web using Kagi's superior search technology.

Prerequisites

A Kagi API key (currently in private beta)
Node.js installed on your system

Installation

Using Smithery CLI (Recommended)

The easiest way to install the Kagi Search MCP is via Smithery:

npx @smithery/cli install kagi-server --client claude

Manual Installation

Clone the repository:

git clone https://github.com/ac3xx/mcp-servers-kagi.git
cd mcp-servers-kagi

Install dependencies:
```
npm install
```
Build the server:
```
npm run build
```

Configuration

Setting up your API Key

Create a .env file in the root directory of the project with your Kagi API key:

KAGI_API_KEY=your_api_key_here

Configuring Claude Desktop

To use with Claude Desktop, add the server configuration to your Claude Desktop config file:

On MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json On Windows: %APPDATA%/Claude/claude_desktop_config.json

Add the following configuration:

{
  "mcpServers": {
    "kagi-server": {
      "command": "/path/to/kagi-server/build/index.js",
      "env": {
        "KAGI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Replace /path/to/kagi-server/build/index.js with the actual path to the built server on your system.

Usage

Once configured, you can use the Kagi search functionality by asking your AI assistant questions that require web searches. For example:

"Search for the latest news about quantum computing"
"Find information about climate change initiatives in Europe"
"Look up recipes for vegan lasagna"

The AI assistant will use the kagi_search tool to retrieve results and provide you with relevant information.

Debugging

Since MCP servers communicate over stdio, debugging can be challenging. You can use the MCP Inspector for debugging:

npm run inspector

This will provide a URL to access debugging tools in your browser.

Limitations

Currently only the kagi_search tool is implemented
Kagi API is in private beta, so you'll need to request access
Planned features like summarization and FastGPT are not yet available