WikiTeam is a GPL-3.0-licensed set of Python tools for downloading and preserving MediaWiki wikis, built for archivists, wiki operators, and volunteers who want to capture a wiki before it disappears.

What it is

WikiTeam is a collection of command-line Python tools for archiving wikis. It works against MediaWiki, the engine behind Wikipedia, and the project intends to expand to other wiki engines over time. It operates as the Archive Team subcommittee on wikis, and it was founded and originally developed by Emilio J. Rodríguez-Posada, a Wikipedia veteran editor and amateur archivist. As of 2026, WikiTeam has preserved more than 600,000 wikis, several wikifarms, regular Wikipedia dumps, and 34 TB of Wikimedia Commons images.

The concrete problem is disappearance. Thousands of wikis exist on the internet, most of them under free licenses, and every day some of them stop being publicly available and are lost forever because nobody kept a backup. Millions of people download media files as a form of distributed backup, but wikis rarely get the same treatment. WikiTeam replaces the absence of that discipline: it stands in for the ad hoc, manual saving of individual pages, or for no preservation attempt at all. The main tool, dumpgenerator.py, replaces page-by-page copying with a single command that walks a wiki's API and writes complete XML page histories and media files to disk.

Key capabilities

• dumpgenerator.py downloads complete XML histories and images from any MediaWiki wiki with python dumpgenerator.py http://wiki.domain.org --xml --images.
• Scope is adjustable: --xml captures histories only, --images captures media only, and --xml --curonly captures just the current version of every page.
• When the script cannot discover the API or index.php paths itself, they can be supplied explicitly with --api= and --index=.
• Interrupted downloads resume with --resume --path=/path/to/incomplete-dump.
• wikipediadownloader.py downloads Wikimedia XML dumps covering Wikipedia, Wikibooks, Wikinews, and other Wikimedia projects.
• Wikimedia Commons image tarballs are published on the Internet Archive, where reseeding the existing torrents is more useful than regenerating them with the script.
• The test suite runs through the tox command, which requires version 1.6.

Who uses it and how

• Volunteer archivists operating under Archive Team, who collectively preserve collections of wikis rather than single sites.
• Individual wiki operators and community members who want a local backup of a live wiki before it is shut down or goes offline.
• Large-scale preservation efforts that have accumulated more than 600,000 wikis, with the resulting backups catalogued in the project wiki.
• People who need Wikimedia XML dumps and prefer the regular dump pipeline that wikipediadownloader.py provides.
• Contributors who send suggestions, report bugs through GitHub issues, write documentation on the wiki, or ask questions on the wikiteam-discuss mailing list.

Getting started

The tools require Python 2.7, and dependencies are installed with pip install --upgrade -r requirements.txt, or pip install --user --upgrade -r requirements.txt when permissions are insufficient. After that, a wiki is archived by running python dumpgenerator.py http://wiki.domain.org --xml --images.

How it compares

This registry's facts name no comparable wiki downloader or preservation tool, so WikiTeam stands alone here rather than being measured against direct alternatives.

When to use it — and when not to

The clearest trade-off is the hard Python 2.7 requirement, an end-of-life interpreter that means a self-hoster has to maintain a dedicated legacy environment just to run the tools. The repository also carries 171 open issues and a README that is deliberately quick rather than complete, pointing users to the wiki tutorial for anything beyond the most common flags. Support for wiki engines other than MediaWiki is a stated goal rather than a delivered feature, so anyone archiving a non-MediaWiki wiki should not expect the current tools to cover it.

$ tox
py27 runtests: commands[0] | nosetests --nocapture --nologcapture
Checking http://wiki.annotation.jp/api.php
Trying to parse かずさアノテーション - ソーシャル・ゲノム・アノテーション.jpg from API
Retrieving image filenames
.    Found 266 images
.
-------------------------------------------
Ran 1 test in 2.253s

OK
_________________ summary _________________
  py27: commands succeeded
  congratulations :)
$