مدیاویکی:Titleblacklist

Title blacklist
One regex per line, compiled to /^PATTERN$/usi — anchored to the WHOLE title and
case-insensitive, which is why every entry is wrapped in .* to mean "contains".
Underscores are normalised to spaces. Usernames are matched as "User:Name", with
the prefix included, so a pattern anchored at the start must allow for it.
Flags go in <angle brackets> after the regex, pipe-separated:
<autoconfirmed> rule applies to NON-autoconfirmed users only
<newaccountonly> gates account creation only, never page titles
<moveonly> gates moves only <noedit> also blocks editing existing pages
<casesensitive> drop the /i <errmsg=key> use MediaWiki:key for the error
Because $wgTitleBlacklistUsernameSources = '*', an unflagged entry gates new account
names as well as titles. Blocks creation / move / upload / account creation — NOT
edits to existing pages (<noedit> is deliberately unused).
Exceptions go in MediaWiki:Titlewhitelist, which is applied AFTER this list and
overrides it. Whitelist a legitimate title rather than weakening a rule here.
Docs: https://www.mediawiki.org/wiki/Extension:TitleBlacklist

--- URLs in titles: always spam, nobody is exempt ---

.*https?:\/\/.* .*www\.[a-z0-9-]+\.[a-z]{2,}.*

--- Keyboard mashing: 10+ identical characters in a row ---

.*(.)\1{9,}.*

--- Broad rules: real subjects can trip these (Booking.com, an article about a
--- 9-digit number), so they bite new accounts only. Established editors are
--- trusted here for the same reason they hold skipcaptcha.

.*[a-z0-9-]+\.(com|net|org|info|biz|xyz|top|shop|online|site|club)\b.* <autoconfirmed> .*\d{9,}.* <autoconfirmed> .*@[a-z0-9-]+\.[a-z]{2,}.* <autoconfirmed>

--- Usernames impersonating staff ---
Substring match, so this also catches names that merely CONTAIN one of these
(e.g. "Badminton" contains "admin"). Whitelist collateral in MediaWiki:Titlewhitelist
rather than adding word boundaries — \b is unreliable across scripts.