PlugWiki
ddgs logo

DDGS MCP Server (deedy5): Setup for Claude Code, Cursor and Other Apps

CommunityLocal
GetPyPIddgsdeedy5/ddgs
Tools
6
6 read · 0 change
Works in
12
apps
Risk
Low
read only
Auth
None

Gives Claude, Cursor and other assistants free web, news, image, video and book search, plus a tool that fetches a page.

DDGS is a metasearch library that asks several search engines at once, including DuckDuckGo, Bing, Brave, Google and Wikipedia. Its optional MCP server hands those searches to your assistant. It also fetches a web page and returns its content.

It runs on your machine and needs no account or API key. The search library is the main project, so the MCP server is an add-on you enable with the mcp extra.

Results come from scraping search engines, so quality and availability can vary.

Good for

  • Anyone who wants web search in an assistant without paying for a search API
  • Developers building research or fact-checking workflows in Claude Code or Cursor
  • Privacy-minded users who want searches sent from their own machine

Works in

12 should work

Tap an app to see how to set it up and any known problems.

Install ddgs MCP

Claude Code

  1. 1

    Copy this prompt.

    Add the ddgs MCP server to Claude Code. Run this command:

claude mcp add ddgs -- uvx --from 'ddgs[mcp]' ddgs mcp

Then check that it loaded and tell me if I need to sign in or restart.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires uv (Python).

Read-only mode: There is no read-only switch, and none is needed. All six tools only search the web or fetch a page.

Install uv first. The command uses the mcp extra of the ddgs package because the plain package does not include the MCP server, and ends in mcp to start it. It works without an account or API key. Add the proxy option from the README after mcp to route searches through a proxy.

Built from the package data we hold. Not yet tested by us.

Tools

  • Searches the web and returns titles, links and snippets

    Runs a web search across several engines and returns up to 10 results by default. You can set region, safe search, a time range, a page number or a specific engine.

    Try: “Search the web for the latest Python 3.14 release notes from the past week.”

    Network · search_text

  • Searches the web for images

    Returns image results for a query, with the same region and safe-search options.

    Try: “Find images of the Golden Gate Bridge at sunset.”

    Network · search_images

  • Searches recent news

    Returns news articles for a query, which you can narrow by time range.

    Try: “What is the news today about electric vehicle tax credits?”

    Network · search_news

  • Searches for videos

    Returns video results for a query.

    Try: “Find videos explaining how transformers work.”

    Network · search_videos

  • Searches for books

    Returns book results for a query. The README lists Anna's Archive as the only book source, so check that downloading what it finds is legal where you live.

    Try: “Find books about the history of cryptography.”

    Network · search_books

  • Fetches a web page and returns its content

    Downloads the page at a URL and returns it as Markdown, plain text, rich text, raw HTML or raw bytes.

    Try: “Fetch this page and give me the content as Markdown.”

    Network · extract_content

Read from the source code

Permissions

low risk · 3 things to watch

Low risk

All six tools only read from the web, and none can change or delete anything on your machine. Search terms and any address you ask it to fetch are sent to outside sites, and it can fetch addresses that only your network can reach.

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • It can fetch any address your machine can reach

    The page tool takes any URL, including addresses on your private network. Do not ask it to open links you do not trust. github.com

  • Your searches leave your machine

    The server scrapes search engines directly from your computer, so your queries and IP address go to those engines unless you set a proxy. github.com

  • Book search uses Anna's Archive

    The README lists Anna's Archive as the only book backend. Copyright rules for what you find there depend on your country. github.com

Security advisories

No published security advisories as of Nov 1, 2026.

Technical details
Turning on read-onlyNo switch is documented, and none is needed because every tool only searches or fetches.
Talks toSearch engines such as DuckDuckGo, Bing, Brave, Google and Wikipedia. Any web address you ask it to fetch
Sign-in and keysNo account or API key. An optional DDGS_PROXY environment variable sends searches through a proxy.

Questions

Is this an official DuckDuckGo server?

No. It is a community project by deedy5 and is not made by DuckDuckGo or any other search engine. It queries DuckDuckGo, Bing, Google and others by scraping their public results pages.

Does the DDGS MCP server need an API key?

No. It works without an account or key. It scrapes public search engines from your own machine.

Is it free?

Yes. It is free and MIT licensed.

Why does the install command mention an extra?

The MCP server is an optional part of the package. Plain ddgs does not include it, so the command asks for the mcp extra and then starts it with the mcp command.

Which search engines does it use?

For text search the README lists Bing, Brave, DuckDuckGo, Google, Grokipedia, Mojeek, Startpage, Yandex, Yahoo and Wikipedia. The default picks one automatically.

Can I send searches through a proxy or Tor?

Yes. The README shows a proxy option after the mcp command, and the source also reads a DDGS_PROXY environment variable.

Does it run locally?

Yes. Your assistant starts it on your machine over standard input, and it contacts search engines directly.

Information

updated Nov 2, 2026

Publisher
Community
Package
ddgs
Registry
PyPI
Transport
stdio
Auth
None
License
MIT
Last commit
Oct 23, 2026
Open issues
18

GitHubChangelogIssues

Sources

Docs, issue trackers and discussions this page is built from. Issues, quotes and findings above link to their own source.

Show all 3
  1. ddgs README · github.com
  2. ddgs MCP server source · github.com
  3. ddgs on PyPI · pypi.org

Facts from the official MCP registry, GitHub, npm, PyPI and the Docker catalog. Updated Nov 2, 2026.