spences10/mcp-omnisearch

πŸ” A Model Context Protocol (MCP) server providing unified access to multiple search engines (Tavily, Brave, Kagi, Exa), AI tools (Kagi FastGPT, Exa, Linkup), and content extraction services (Firecrawl, Tavily, Kagi). Includes GitHub search. All through a single interface.

What it solves

This project provides a unified interface for AI models to access multiple search and data extraction tools. Instead of installing separate servers for every search provider, it consolidates multiple services like Tavily, Brave, Kagi, Exa AI, GitHub, Linkup, and Firecrawl into a single Model Context Protocol (MCP) server.

How it works

It acts as a bridge between an MCP-compatible AI client (like Claude) and various external APIs. It exposes four consolidated tools to the AI:

  • web_search: Performs web searches across different providers.
  • ai_search: Retrieves sourced AI-generated answers from specific providers.
  • ai_search: Retrieves sourced AI-generated answers from specific providers.
  • github_search: Searches for code, repositories, or users on GitHub.
  • web_extract: Extracts, crawls, or scrapes content from the web using providers like Firecrawl or Tavily.

Who it’s for

Developers and AI users who want to give their LLMs the ability to search the web, search GitHub, and extract web content without managing multiple separate MCP servers.

Highlights

  • Unified Access: Combines seven different search and extraction providers into one server.
  • Flexible Provider Selection: Allows users to configure which providers to use based on available API keys.
  • Crawl and Extract: Includes capabilities for scraping and summarizing content via tools like Firecrawl.
  • GitHub Integration: Built-in support for searching GitHub code and repositories.

κ΄€λ ¨

  • ν”„λ‘œμ νŠΈ
    anysearch-ai/anysearch-mcp-server

    AI μ—μ΄μ „νŠΈκ°€ ν†΅ν•©λœ μ‹€μ‹œκ°„ μ›Ή 검색, 수직 도메인 검색, URLμ—μ„œμ˜ Markdown μ½˜ν…μΈ  μΆ”μΆœμ„ κ°€λŠ₯ν•˜κ²Œ ν•˜λŠ” MCP μ„œλ²„μž…λ‹ˆλ‹€.

  • ν”„λ‘œμ νŠΈ
    1mcp-app/agent

    μ—¬λŸ¬ MCP μ„œλ²„λ₯Ό ν•˜λ‚˜μ˜ ν”„λ‘œμ„ΈμŠ€λ‘œ ν†΅ν•©ν•˜λŠ” 톡합 λŸ°νƒ€μž„μœΌλ‘œ, μ—μ΄μ „νŠΈμ˜ μ»¨ν…μŠ€νŠΈ μ‚°λ§Œν•¨μ„ λ°©μ§€ν•˜λŠ” 점진적인 도ꡬ 탐색 μ›Œν¬ν”Œλ‘œμš°λ₯Ό μ œκ³΅ν•©λ‹ˆλ‹€.

  • ν”„λ‘œμ νŠΈ
    samanhappy/mcphub

    Model Context Protocol (MCP) μ„œλ²„λ₯Ό ν†΅ν•©μ μœΌλ‘œ κ΄€λ¦¬ν•˜λŠ” ν—ˆλΈŒλ‘œ, 이λ₯Ό 슀트리밍 κ°€λŠ₯ν•œ HTTP μ—”λ“œν¬μΈνŠΈλ‘œ μ •λ¦¬ν•˜κ³  쀑앙 집쀑식 λΌμš°νŒ… 및 AI 기반 도ꡬ 탐색을 μ œκ³΅ν•©λ‹ˆλ‹€.

  • ν”„λ‘œμ νŠΈ
    blazickjp/arxiv-mcp-server

    AI μ–΄μ‹œμŠ€ν„΄νŠΈκ°€ arXivλ₯Ό κ²€μƒ‰ν•˜κ³ , 논문을 λ‹€μš΄λ‘œλ“œν•˜λ©°, 전체 ν…μŠ€νŠΈλ₯Ό 읽고, ν•™μˆ  연ꡬλ₯Ό μœ„ν•œ 인용 κ·Έλž˜ν”„λ₯Ό 뢄석할 수 μžˆλ„λ‘ ν•˜λŠ” MCP μ„œλ²„μž…λ‹ˆλ‹€.

  • ν”„λ‘œμ νŠΈ
    brightdata/brightdata-mcp

    봇 및 CAPTCHA 우회 κΈ°λŠ₯이 λ‚΄μž₯λ˜μ–΄ 검색, μŠ€ν¬λ ˆμ΄ν•‘ 및 κ΅¬μ‘°ν™”λœ 데이터 μΆ”μΆœμ„ μœ„ν•œ 69개의 도ꡬλ₯Ό μ œκ³΅ν•˜λ©°, AI μ—μ΄μ „νŠΈμ—κ²Œ μ‹€μ‹œκ°„ μ›Ή μ•‘μ„ΈμŠ€λ₯Ό μ œκ³΅ν•˜λŠ” MCP μ„œλ²„μž…λ‹ˆλ‹€.