HTML analysis
ID: cia-2010-covert-communication-websites/html-analysis
CIA 2010 covert communication websites HTML analysis by
Ciro Santilli 37 Updated 2025-06-02 +Created 2025-05-07
The HTML from the index page of Wayback Machine were:
- dumped at: github.com/cirosantilli/media/tree/master/cia-2010-covert-communication-websites/html
- downloaded with: github.com/cirosantilli/media/tree/master/cia-2010-covert-communication-websites/download-html.sh. Note that there were many supurious errors notably:we just ran it multiple times until all errors were gone.
OpenSSL SSL_connect: SSL_ERROR_SYSCALL in connection to web.archive.org:443
The best way to analyse the HTML is to grap our dumps from: github.com/cirosantilli/cia-2010-websites-dump.
Some possibly interesting searches include:
Some of the HTML files contain conditional comments e.g. web.archive.org/web/20091023041107/http://aquaswimming.com/ contains:
<!--[if IE 6]> <link href="swimstyleie6.css" rel="stylesheet" type="text/css"> <![endif]-->
Varios of the non-English websites seem to have comments translating the content e.g.:This feels like it could be the translation helping the technical webdev team know what is what.
./noticiasmusica.net/20101230165001/index.html:<h2>Alguns dos Melhores Sites Nacionais</h2><!--some of the best national sites (in music)-->
Many of the RSS frame pages use:which is a weird HTML tag that would lead all links to open on new tabs, e.g. web.archive.org/web/20110202124411/http://thecricketfan.com/home.html.
<base target="_blank" />
Various websites have pages with .php extension. It feels likely that all websites were written in PHP.
Some sites use a
feeds.php
for the feeds, e.g. http://www.absolutebearing.net//absolutebearing_feeds/feeds.php?src=http%3A%2F%2Ffeeds2.feedburner.com%2FOceanyachtsinfo&desc=1Some URLs existed both in HTML and .php extension, or were converted at some point:
allworldstatistics.com/20110207151941/comprehensivesources.html
allworldstatistics.com/20130818155225/comprehensivesources.php
A few of the PHP urls have weird IDs in them like we wonder what they mean.
omktf
, juqwt
and qlaqft
:./middle-east-newstoday.com/20100829004127/omktf/uirl.php?ok=461128
./newsandsportscentral.com/20100327130237/juqwt/eubcek.php?pe=747155
./pondernews.net/20100826031745/lldwg/qlaqft.php?fc=281298
A few separate websites have an archive with the same It is unclear what it means. All of them contain something like:so looks like an archival artifact only.
pid
parameter:fightwithoutrules.com/20131220205811/?pid=2POQ7BC1G/index.html
half-court.net/20131223165013/?pid=2POQ7BC1G/index.html
health-men-today.com/20131223002237/?pid=2POQ7BC1G/index.html
intlnewsdaily.com/20131221121441/?pid=2POQ7BC1G/index.html
intoworldnews.com/20131217193621/?pid=2POQ7BC1G/index.html
<html>
<head>
<meta name="robots" content="noarchive" />
<meta name="googlebot" content="nosnippet" />
</head>
<body>
<div align=center>
<h3>Error. Page cannot be displayed. Please contact your service provider for more details. (11)</h3>
</div>
</body>
</html>
The following two websites have a
feeds.php
system for their RSS:./mydailynewsreport.com/20110211111053/myrss/feeds.php?src=http:/www.refahemelli.com/pashto/news/rss.php&chan=y&desc=1&targ=y&utf=y
./magneticfieldnews.com/20110208063545/magneticfeeds/feeds.php?src=http:/www.bbc.co.uk/pashto/index.xml&chan=y&desc=1&targ=y&utf=y
Some of the HTML uses attributes without quotes, which is legal, but very unusual nowadays:
soldiersofsouthasia.com/20110207203705/home.htm: <a href=http://www.rss-to-javascript.com
We can try to search for any link leaks by listing all domains linked to with:The first thing that shows up is that there are some IPs linked to directly! But they seem to be the direct IPs of legitimate websites, we are not sure why IPs were used rather than domain names:
git grep --no-color -I -h --no-line -o 'https?://[^/">?]+[/">?]' | sed -r 's/.$//' | sort | uniq -c | sort -nk1
- 69.167.160.171 at web.archive.org/web/20110208053653/http://sa-michigan.com/ to web.archive.org/web/20100304122019/http://69.167.160.171/ marked with image "fantasyplayers.com", a legit website called Fantasy Players Network
- 69.94.11.53 at web.archive.org/web/20101229193800/http://newsresolution.net/ titled "International Tribunal for Rwanda" to web.archive.org/web/20101229193800/http://69.94.11.53/default.htm
- 74.125.77.132 mynepalnews.com Webalizer
- 194.165.154.66/index.php web.archive.org/web/20110129161937/http://icwb-news.com/ MiddleEast links to 194.165.154.66/index.php but that is an actual page: web.archive.org/web/20110529142501/http://194.165.154.66/index.php
- 200.55.6.87 at web.archive.org/web/20110128170204/http://noticiasdelmundolatino.com/ after clicking "Maps" tab entitled "Mapas en la red" to web.archive.org/web/20100329150648/http://200.55.6.87/es/index.htm
- 213.97.154.118 at web.archive.org/web/20120429042725/http://montanismoaventura.com/ entitled "Mallorca Verde" to web.archive.org/web/20120430191214/http://213.97.154.118/mallorcaverde/ The target is a bit weird and almost empty.
- 216.218.196.146 at entitled "AskTheDr.com" to web.archive.org/web/20070303080403/http://216.218.196.146/askthedr/index.htm
We can also get the full line for each with sorted by least common domains with the slow:
git grep --no-color -I -h --no-line -o 'https?://[^/">?]+[/">?]' | sed -r 's/.$//' | sort | uniq -c | sort -nk1 | awk '{if ($1 < 10) print $2}' | xargs -I{} git --no-pager grep -h --no-line -o '{}.*<' | tee tmp.log
We can search for all IP-like strings with:
git grep '[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\b'
New to topics? Read the docs here!