Search MCP Server

Unified MCP search tools for AI agents — abstracts providers, caching, reranking, and rate limits behind a few simple tools.

Tools

Related MCP server: Crawl4AI+SearXNG MCP Server

Quick Start

git clone <repo-url> search-mcp
cd search-mcp
npm install
cp .env.example .env
npm run build
npm start

Configuration

Copy .env.example to .env and set your API keys. No keys are required — DuckDuckGo and Bing work out of the box.

Tool Reference

search

search({
  query: string,
  intent?: "web" | "docs" | "github" | "news",
  freshness?: "any" | "day" | "week" | "month",
  max_results?: number,       // default 10, max 30
  include_content?: boolean   // fetch page content as markdown
})

Returns merged results from the first 2 healthy providers (parallel), deduplicated and reranked by relevance.

github_search

github_search({
  query: string,
  type?: "repositories" | "code" | "issues" | "users",
  language?: string,
  stars?: string,             // e.g. ">1000", "500..5000"
  per_page?: number,          // default 15, max 100
  page?: number
})

Rate limit: 60 req/hr without token, 5000 req/hr with GITHUB_TOKEN.

gitlab_search

gitlab_search({
  query: string,
  scope?: "projects" | "issues" | "merge_requests" | "blobs",
  per_page?: number,          // default 10, max 100
  page?: number
})

Requires GITLAB_TOKEN with read_api scope.

status

status()

Returns provider health, cache stats, budget state, uptime.

Architecture

Pipeline

Core pipeline: Normalize → Budget Check → Cache → Router (parallel 2) → Rerank → Cache → Respond

Full docs:

• Architecture

• Providers & Fallback

• Caching

• Reranking

• Fetch Layer

• Budget System

• Configuration

• Roadmap

License

MIT