MCP Retrieval

io.github.Role1776/mcp-retrieval

Local Go MCP server: DuckDuckGo/Bing web + image search, readability scraping to Markdown.

Connect this server

Add this to your MCP client config — claude_desktop_config.json (Claude Desktop) or .cursor/mcp.json (Cursor).Suits: Claude Desktop · Cursor.

{
  "mcpServers": {
    "mcp-retrieval": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/role1776/mcp-retrieval:0.1.10"
      ],
      "env": {
        "MCP_TRANSPORT": "stdio",
        "MCP_NAME": "mcp-retrieval",
        "MCP_PATH": "/mcp",
        "SERVER_PORT": "8080",
        "SERVER_READ_TIMEOUT": "60s",
        "SERVER_WRITE_TIMEOUT": "60s",
        "LOG_MODE": "local",
        "MAX_IDLE_CONNS_PER_HOST": "100",
        "MAX_QUERIES": "10",
        "DEFAULT_RESULTS": "5",
        "MAX_RESULTS": "20",
        "DEFAULT_IMAGES": "5",
        "MAX_IMAGES": "10",
        "DEFAULT_TIMEOUT_MS": "5000",
        "MIN_TIMEOUT_MS": "1000",
        "MAX_TIMEOUT_MS": "10000",
        "DEFAULT_DOCUMENT_CHARS": "20000",
        "MAX_DOCUMENT_CHARS": "20000",
        "PROXY_LOGIN": "<your PROXY_LOGIN>",
        "PROXY_PASSWORD": "<your PROXY_PASSWORD>"
      }
    }
  }
}

Environment variables

MCP_TRANSPORTstdio (client spawns the process) or http (streamable HTTP listener).
MCP_NAMEServer name reported in the MCP initialize handshake.
MCP_PATHHTTP path the MCP endpoint is served on. Used only when MCP_TRANSPORT=http.
SERVER_PORTTCP port to listen on. Used only when MCP_TRANSPORT=http.
SERVER_READ_TIMEOUTHTTP read timeout as a Go duration. Used only when MCP_TRANSPORT=http.
SERVER_WRITE_TIMEOUTHTTP write timeout as a Go duration. Used only when MCP_TRANSPORT=http.
LOG_MODElocal (text, debug) or prod (JSON, info). Logs go to stderr.
MAX_IDLE_CONNS_PER_HOSTIdle connections the outbound HTTP client keeps per host.
MAX_QUERIESMaximum queries or URLs accepted in one call.
DEFAULT_RESULTSSnippets returned per query when the call does not ask for a count.
MAX_RESULTSUpper bound for snippets per query.
DEFAULT_IMAGESImages returned per query when the call does not ask for a count.
MAX_IMAGESUpper bound for images per query.
DEFAULT_TIMEOUT_MSPer-call timeout in milliseconds when the call does not request one.
MIN_TIMEOUT_MSLowest timeout a call may request, in milliseconds.
MAX_TIMEOUT_MSHighest timeout a call may request, in milliseconds.
DEFAULT_DOCUMENT_CHARSCharacters kept per scraped page when the call does not ask for a limit.
MAX_DOCUMENT_CHARSUpper bound for characters kept per scraped page.
PROXY_LOGINsecretOutbound proxy username. Required when PROXY_HOST is set.
PROXY_PASSWORDsecretOutbound proxy password. Required when PROXY_HOST is set.

A starting point — check the repository for exact arguments and setup. Restart your client after editing the config.

Category: Search & Web
Version: 0.1.10
Registry: active
Published: 10/24/2026

Browse all Search & Web →

PyreCrawl

SanggonBoy

Web scraping + research toolkit for AI agents: scrape, crawl, monitor URLs with anti-bot bypass.

Search & WebAdded 1 month ago

Search Console MCP

saurabhsharma2u

Google Search Console + Bing + GA4 + AdSense — unified SEO intelligence for AI agents.

Search & WebAdded 2 weeks ago

ScrapeBadger

scrape-badger

Twitter/X, Google Search, Yelp, GitHub repositories and AutoScout24 listings for AI agents.

Search & WebAdded 3 days ago

io.github.scrapeless-ai/scrapeless-mcp-server

scrapeless-ai

Web search, browser automation, scraping, crawling and CAPTCHA solving for AI agents.

Search & WebAdded 2 months ago

SerpApi

serpapi

Official SerpApi MCP server for Google, Bing, and other search engines.

Search & WebAdded 6 days ago

Serper.live

serper-live

Live Google search results, domain positions and open markets via the Serper.live SERP API.

Search & WebAdded 6 days ago

Add the badge to your site

Show visitors that MCP Retrieval is listed here — it’s also the link we check for a free listing.

HTML — for your website

<a href="https://serverdex.io/server/io-github-role1776-mcp-retrieval?utm_source=badge"><img src="https://serverdex.io/brand/badge.svg" alt="MCP Retrieval — listed on Serverdex" width="123" height="40" loading="lazy"></a>

Recommended. A dofollow link from your own domain is the one that counts for search engines, and it's the link we check for a free listing.

Markdown — for your README

[![MCP Retrieval — listed on Serverdex](https://serverdex.io/brand/badge.svg)](https://serverdex.io/server/io-github-role1776-mcp-retrieval?utm_source=badge)

Nice for visibility, but GitHub marks README links nofollow, so it doesn't pass authority. Use the HTML badge on your site as well.