MCP 服务器

Deep Search MCP Server

Provides comprehensive search capabilities including web search, content extraction, news search, academic search, and AI-powered multi-source research. Enables natural language access to web content and research through a production-ready MCP server.

README

Deep Search MCP Server

A production-ready Model Context Protocol (MCP) server providing comprehensive deep search capabilities including web search, content extraction, and AI-powered research.

Features

Web Search - Search the web for any topic with comprehensive results
Content Extraction - Extract clean, readable content from any URL
Deep Research - Multi-source research with summaries and key findings
News Search - Find recent news articles on any topic
Academic Search - Discover academic papers and research content

Quick Start

Connect to Your AI Assistant

Add this configuration to your MCP client (e.g., Cursor, Claude Desktop):

{
  "mcpServers": {
    "deep-search": {
      "url": "https://deep-search-mcp.vercel.app/api/mcp"
    }
  }
}

For Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "deep-search": {
      "url": "https://deep-search-mcp.vercel.app/api/mcp"
    }
  }
}

Available Tools

Tool	Description	Parameters
`web_search`	Search the web for information	`query` (required), `maxResults` (optional, 1-20)
`extract_content`	Extract content from a URL	`url` (required)
`deep_research`	Comprehensive multi-source research	`query` (required), `depth` (optional, 1-5)
`news_search`	Search for news articles	`query` (required), `maxResults` (optional, 1-20)
`academic_search`	Search academic/research content	`query` (required), `maxResults` (optional, 1-20)

Self-Hosting

Prerequisites

Node.js 20+
npm, yarn, or pnpm

Installation

# Clone the repository
git clone https://github.com/LikhonSheikh404/deep-search-mcp.git
cd deep-search-mcp

# Install dependencies
npm install

# Copy environment file
cp .env.example .env

# Start development server
npm run dev

Deploy to Vercel

Fork this repository
Import to Vercel
Deploy!

Or use the CLI:

# Install Vercel CLI
npm i -g vercel

# Deploy
vercel

Testing with MCP Inspector

npx @modelcontextprotocol/inspector@latest http://localhost:3000

Then open http://127.0.0.1:6274 to test your tools.

API Endpoints

Endpoint	Description
`GET/POST /api/mcp`	MCP server endpoint
`GET /.well-known/oauth-protected-resource`	OAuth metadata (for authenticated setups)

Technology Stack

Framework: Next.js 15 (App Router)
Language: TypeScript
MCP SDK: mcp-handler
Deployment: Vercel (Serverless + Fluid Compute)
Transport: Streamable HTTP (efficient, no persistent connections)

Architecture

src/
├── app/
│   ├── api/
│   │   └── [transport]/
│   │       └── route.ts      # MCP server endpoint
│   ├── .well-known/
│   │   └── oauth-protected-resource/
│   │       └── route.ts      # OAuth metadata
│   ├── page.tsx              # Landing page
│   └── layout.tsx            # Root layout
├── lib/
│   └── search-engine.ts      # Search functionality
└── types/
    └── mcp.ts                # Type definitions