Kagimcp

🚀 Kagi MCP Server

This is a server that provides specific services. It can be integrated with certain tools to perform search and summarization operations.

🚀 Quick Start

✨ Features

• Integrate with search API to provide search and summarization functions.
• Support installation on multiple operating systems.
• Allow for advanced configuration of log levels and summarizer engines.

📦 Installation

⚠️ Important Note

Unless you are only using non - search tools, make sure you have access to the search API. Currently, this feature is in a closed - beta test and requires application. If necessary, please contact support@kagi.com.

First, install uv.

MacOS/Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Installation via Smithery

Alternatively, you can install Kagi for Claude Desktop via Smithery:

npx -y @smithery/cli install kagimcp --client claude

Configuration on Claude Desktop

// claude_desktop_config.json
// The location can be found as follows:
// Button Menu -> File -> Settings -> Developer -> Edit Configuration
{
  "mcpServers": {
    "kagi": {
      "command": "uvx",
      "args": ["kagimcp"],
      "env": {
        "KAGI_API_KEY": "YOUR_API_KEY_HERE",
        "KAGI_SUMMARIZER_ENGINE": "YOUR_ENGINE_CHOICE_HERE" // If the environment variable is not set, the default engine is "cecil"
      }
    }
  }
}

💻 Usage Examples

Basic Usage

• Sample Queries: Here are examples of using the tool:
  • Search: Kagi.search("test")
  • Summarize: Kagi.summarize("test", "english")
• Debug Command: Run the following command in the terminal to start the MCP server:

uvx kagimcp -- KAGI_API_KEY=your_api_key_here

Advanced Usage

• The log level can be adjusted via the FASTMCP_LOG_LEVEL environment variable (e.g., FASTMCP_LOG_LEVEL="ERROR").
  • Related issue: Here
• The summarizer engine can be customized using the KAGI_SUMMARIZER_ENGINE environment variable (e.g., KAGI_SUMMARIZER_ENGINE="daphne").
  • For information on different summarization engines, refer to here