duckduckgo

DuckDuckGo MCPサーバー — AIエージェント向けの軽量で認証不要なウェブ検索ツール。シンプルなMCP互換APIを介して構造化された検索結果(タイトル、URL、スニペット)を提供し、LLMワークフローへの高速な統合に最適化されています。

ドキュメント

🚀 DuckDuckGo MCP Server

A lightweight MCP-compatible tool that enables AI agents to perform real-time web searches and retrieve structured results.

🔍 What It Does

This tool allows AI systems to search the web and get:

  • Relevant page titles
  • Direct URLs
  • Concise snippets for context Designed for seamless integration into agent workflows and LLM pipelines.

✨ Key Features

  • 🌐 Real-time web search
  • 📦 Structured, LLM-friendly responses
  • ⚡ Fast and minimal interface
  • 🔓 No authentication required
  • 🤖 MCP-compatible tool schema

🧠 Why Use This

  • Add live web knowledge to AI agents
  • Improve answer accuracy with fresh data
  • Enable research and browsing capabilities
  • Simple drop-in tool for MCP ecosystems

🛠️ Available Tool

search

Search the web and return relevant results.

Input

json{
  "query": "string"
}

Output

json{
  "results": [
    {
      "title": "string",
      "url": "string",
      "snippet": "string"
    }
  ]
}

🔌 How to Use

1. Discover tools

GET /tools

Returns available tools and their schemas.


2. Execute search

POST /tools/search

Example

json{
  "query": "latest AI news"
}

⚡ Example Response

json{
  "results": [
    {
      "title": "Latest AI Developments",
      "url": "https://example.com",
      "snippet": "Recent advancements in artificial intelligence..."
    }
  ]
}

🎯 Ideal For

  • AI agents and copilots
  • Research assistants
  • RAG pipelines
  • Developer tools and MCP marketplaces

⚠️ Notes

  • Results are optimized for relevance and readability
  • Best used where fresh external information is required
  • Response format is stable and predictable for LLM usage

💡 Summary

A simple, effective way to give AI systems real-time web search capability through a clean MCP interface.