SCANTRAP: Protecting Content Management Systems from
Vulnerability Scanners with Cyber Deception and Obfuscation
Daniel Reti
1 a
, Karina Elzer
1 b
and Hans Dieter Schotten
1,2 c
1
German Research Center for Artificial Intelligence, Kaiserslautern, Germany
2
Technische Universit
¨
at Kaiserslautern, Germany
Keywords:
CMS, Penetration Testing, Vulnerability Scanner, Information Security, Cyber Deception, WordPress, Plugin,
Website.
Abstract:
Every attack begins with gathering information about the target. The entry point for network breaches are
often vulnerabilities in internet facing websites, which often rely on an off-the-shelf Content Management
System (CMS). Bot networks and human attackers alike rely on automated scanners to gather information
about the CMS software installed and potential vulnerabilities. To increase the security of websites using a
CMS, it is desirable to make the use of CMS scanners less reliable. The aim of this work is to extend the
current knowledge about cyber deception in regard to CMS. To demonstrate this, a WordPress Plugin called
’SCANTRAP’ was created, which uses simulation and dissimulation in regards to plugins, themes, versions,
and users. We found that the resulting plugin is capable of obfuscating real information and to a certain
extent inject false information to the output of one of the most popular WordPress scanners, WPScan, without
limiting the legitimate functionality of the WordPress installation.
1 INTRODUCTION
Content Management Systems (CMS) have become a
notable part of today’s web, with almost half of the
websites using some sort of CMS. This is due to the
ease of use and free option. Specifically, out of the ex-
isting web CMSs, some popular CMSs have emerged:
WordPress (WP) is the most popular CMS, with a
market share of 63.3%. Shopify, Wix and Squares-
pace, as paid alternatives, together have 12.4% of the
market share. The most popular open-source alterna-
tives, Joomla and Drupal, are used in 4.3% of web-
sites using CMS. (W3Techs, 2022)
With such popularity, however, they also become
an attractive target for attackers. Due to the addi-
tion of third-party extensions, these platforms leave
a large attack surface. In the field of penetration test-
ing, different tools have emerged that reduce the ef-
fort of scanning and exploitation of popular CMSs.
This makes reconnaissance possible with low effort,
enabling the threat from so-called ’script kiddies’ and
bots.
a
https://orcid.org/0000-0001-8071-6188
b
https://orcid.org/0000-0002-3984-779X
c
https://orcid.org/0000-0001-5005-3635
Therefore, it is desirable to increase the effort of
attackers in regards to attacking CMS by using differ-
ent defence strategies. This starts with the attacker’s
capability of scanning CMSs. The approach of this
work is based on cyber deception to disguise valuable
information, by hiding the real and presenting false
information.
As the most popular CMS as of today, we choose
WP as a representative CMS to apply deception mea-
sures for our research. Due to its popularity, almost
100,000 plugins and 25,000 themes exist for WP. This
leaves a large attack surface leading to over 30,000
known vulnerabilities. 92% of which are found in
plugins. (WPScan, 2022) Also, because of its pop-
ularity, we choose WPScan, a WP vulnerability scan-
ner maintained by Automattic, the vendor of word-
press.com, as the attacker’s tool of choice.
This paper presents a systematic approach to de-
ception based defence against CMS scanners. Hereby
a WP plugin to defend against WPScan is created
as a proof of concept. This plugin, called ’SCANT-
RAP’, uses simulation and dissimulation techniques
of the cyber deception domain to make reconnais-
sance harder for attackers. The source code can be
found at https://github.com/dfki-in-sec/SCANTRAP.
Additionally, we give an overview of cyber decep-
Reti, D., Elzer, K. and Schotten, H.
SCANTRAP: Protecting Content Management Systems from Vulnerability Scanners with Cyber Deception and Obfuscation.
DOI: 10.5220/0011667400003405
In Proceedings of the 9th International Conference on Information Systems Security and Privacy (ICISSP 2023), pages 485-492
ISBN: 978-989-758-624-8; ISSN: 2184-4356
Copyright
c
2023 by SCITEPRESS Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
485
tion methods for the web domain, the functionality of
different CMS scanners and a survey of cyber decep-
tion methods used by existing WP security plugins.
Our paper is divided into six sections. The follow-
ing section gives a brief overview of the background
of cyber deception and CMS. The third section de-
scribes our created plugin. In the fourth, we present
our discussion and in section five we examine related
work. Our conclusion is drawn in the final section.
2 CONTENT MANAGEMENT
SYSTEM SECURITY AND
CYBER DECEPTION
In the cyber kill chain, a model describing the stages
of an attack, the first phase is reconnaissance. In
this phase, the attacker will gather information that
will allow clearer decision-making in later steps. One
method of reconnaissance is scanning. This paper’s
focus lies on defending this part of the cyber kill
chain through cyber deception in the domain of CMS,
which will be explained in detail in this section.
2.1 Cyber Deception
In the field of cyber security, the term cyber deception
is generally understood as a defence strategy which
is used to intentionally mislead an attacker. Specif-
ically, by simulating fake or hiding existing targets,
the actions and in-actions of the attacker should be-
come more predictable or take more time. (Yuill et al.,
2007)
In the literature, many deception techniques in
mulitple computer domains have been proposed and
studied (Fraunholz et al., 2018a).
In 1991, Bell and Whaley defined a taxonomy for
deception that contains two main classes (Bell and
Whaley, 1991). Dissimulation describes the hiding of
the existing and simulation refers to faking.
The term ’obfuscation’ is a type of cyber decep-
tion and dissimulation, mainly understood in relation
to code protection. However, the idea is to transform
a structure to make it harder for an attacker to deter-
mine the functionality. (Behera and Bhaskari, 2015)
More concrete examples of cyber deception are
honeypots and honeytokens. Honeypots are fake
computer systems that allow the detection of unusual
interactions. Honeytokens serve the same purpose
but are anything but a computer. (de Barros, 2007)
(de Barros, 2007) (Spitzner, 2003)
Regarding cyber deception for web services, mul-
tiple specific techniques can be implemented. In this
work, the following methods were considered, based
on the work of (Fraunholz et al., 2018b):
Version Trickery is a form of dissimulation that
provides falsified version information. This could
be the version of the HTTP Server, the Operating
System, the CMS installation, plugins or themes.
This makes it harder for the attacker to identify
vulnerable versions or find exploits. In the context
of CMSs, it is possible to obfuscate the version
completely, as described in Section 3.3. For WP,
exisiting plugins can fake the version, see Table 2.
Disallow Injection refers to injecting Disallow en-
tries in robots exclusion file. These entries are
used by web crawlers as well as different scan-
ners, such as CMS Scanners, to find possibly sen-
sitive paths and files. This falls into the category
of simulation and can be used for detection pur-
poses. The WP plugin ’Blackhole for Bad Bots’
(Starr, 2022) is an example for faking entries in
the robots file.
Status Code Tampering refers to manipulating the
HTTP response codes and could be used agaist
scanners relying on this status code to test whether
a certain path exists or not. It was used in this
work to hide and fake the existence of specific plu-
gins and is further described in Section 3.1.
Latency Adaption is a simulation tactic to pre-
vent brute-force attacks and can be thought of as
a veiled rate limiting. The more requests are sent,
the more the response from the server will be de-
layed, which limits the number of requests that the
attacker can send. Rate-limiting brute force login
attacks is a method used by different CMS secu-
rity plugins (Sec. 5).
Virtual Honey Files is a method where invented
resources on the website are presented, with the
goal of creating uncertainty for the attacker. This
could be a fake directory, document, login page,
back end, CMS installation etc. In this work a
similar method was used to include plugin version
numbers as described in 3.1.
Code Obfuscation is used to hide the functionality
of code such as JavaScript. This is used to pre-
vent the stealing of intellectual property. It might
be effective against CMS scanners when specific
patterns in the HTML are analysed to extract in-
formation, but has not been applied in this work.
Cookie Scrambling can be used as either simula-
tion or dissimulation tactic. This is done by either
creating a honey cookie, where tampering with
the cookie can be detected or by creating cook-
ies that need to be set to successfully execute a
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
486
task. An example of this for Wp is described by
a blog post (Talk, 2022), which is explained more
closely in Section 5.
Content Modification can be used for dissimula-
tion and simulation. Example for content modifi-
cation are described in Sections 3.3 and 3.4 where
users and versions are hidden. Furthermore, as
seen in Table 2, it can also be used to fake users
and the version.
2.2 CMS
A Content Management System (CMS) is a way of
managing content, which includes the collection and
usage of the content. This allows for consistency and
less repetition for the user. (Boiko, 2005, p. 72) There
are two types of CMS: enterprise CMS, which con-
tains content relevant to the whole organisation, and
web CMS’s, which focus is on managing content for
websites. (Boiko, 2005, p. 79-82)
2.3 CMS Scanner
CMS scanners are a type of vulnerability scanner.
Mainly, they enumerate a CMS to find as much infor-
mation that can be used to determine vulnerabilities.
There are different open-source scanners. Some scan-
ners work for multiple different CMSs others are for
a specific CMS.
In Table 1, a comparison was made between dif-
ferent CMS scanners and their functionalities. Mod-
ule enumeration depends on the CMS and is gener-
ally done through directory brute forcing. For ex-
ample, in WordPress, it would be plugin and theme
enumeration. With one exception, every scanner has
this functionality since these modules prone to con-
tain vulnerabilities. Version detection shows similar
results. User enumeration is possible for ve out of
the nine scanners. The detection of misconfigurations
can be done by six scanners. URLs refer to the de-
tection of ’interesting’ URLs that are not necessarily
searched for. Exposed files include backup, default,
log and other files. Three scanners can do brute-force
login attacks. Drupwn is the only scanner that has the
integrated possibility of also exploiting found vulner-
abilities, while some of the other scanners might link
to possible vulnerabilities. Finally, WPSeku allows
for static code analysis, which no other scanner of-
fers. (InfosecMatter, 2020)
2.4 WPScan
WordPress is the most popular, free and open-source
CMS on the market. It allows the creation of web-
sites and blogs with high customizability through a
wide range of community-created themes and plug-
ins, which allow for customized style and functional-
ity, respectively.
WPScan is one of the most known and advanced
CMS scanners for WordPress with a large feature set
(Tab. 1). In the work the focus is on its capabilities
of theme, plugin, user and version enumeration and
detection.
3 SCANTRAP WPScan EVADER
For evaluating the ability to evade CMS scanners, we
used WordPress as the CMS and WPScan as the CMS
scanner and created a plugin for WordPress that fo-
cused on the four main scans of WPScan. A the
source code of WPScan is publicly available (WP-
ScanTeam, 2022), it was used to better understand
how the scans are performed and how the information
is gathered.
Although WPScan reveals itself by setting the
HTTP User-Agent field to ’WPScan’, this informa-
tion was not used, as it could be easily changed for
malicious purposes. Instead, cyber deception was
used, as explained in 2.1, to manipulate the scanner
output. Additionally, we partly adapted and modified
ideas from a blog post (Cyberpunk, 2019) on this mat-
ter.
3.1 Plugin Detection
Plugin detection is how WPScan finds existing plu-
gins as well as their versions. The focal method for
finding plugins is based on known locations. The
WordPress structure has a determined folder (’/wp-
content/plugins/’) for plugins. Therefore WPScan
does a directory scan for known plugins in this folder.
If the directory for the plugin exists, which is based
on the response code, then the mandatory readme.txt’
file is requested and the version is parsed from either
the ’Stable Tag’ or ’ChangeLog Section’.
For this work, two methods to evade plugin detec-
tion were used. The first approach is based on simula-
tion. We faked plugins by changing the 404 response
code of requests to non-existing plugins. WPScan de-
termines a plugin as existing, when one out of four re-
sponse codes is detected. The code ”200 OK” means
the request succeeded. ”401 Unauthorized” stands for
unauthenticated access and ”403 Forbidden” means
the user has no access rights. Lastly, ”500 Internal
Server Error” indicates an error on the server’s side.
For responses to plugin folder requests, we chose the
403 response code and for the ’readme.txt’ file we se-
SCANTRAP: Protecting Content Management Systems from Vulnerability Scanners with Cyber Deception and Obfuscation
487
Table 1: Comparing different CMS scanners functionality (InfosecMatter, 2020).
Scanner CMS Module Version User Misconfigs URLs Exposed Login Exploits Code
Enumeration Detection Enumeration Files Attacks Analysis
Droopescan multiple X X - - X X - - -
CMSmap multiple X X X X X X X - -
CMSeek multiple X X X X - X - - -
WPScan WordPress X X X X X X X - -
WPSeku WordPress X - - X - X X - X
JoomScan Joomla X X - X - X - - -
JoomlaVS Joomla X X - X - - - - -
JScanner Joomla - X X - - - - - -
Drupwn Drupal X X X - - X - X -
lected status code 200. Further, we added customised
text to the response body to add a stable tag for ver-
sion detection. As an addition, based on honeytokens,
we added simple logging when fake plugin paths are
requested. This extends deception as a defence to also
be able to detect unusual behaviour. An alternative to
logging could have been an email notification.
The second method of hiding the existing plugins,
is a method of dissimulation. The plugins are hidden
by responding to requests to the plugin folder with a
’404 Not Found’ response. This is possible because
WPScan will request files only when the folders are
found. It still allows WordPress to use the contents of
the folders, therefore not restricting the functionality
of the plugins.
Figure 1 shows the sequence of request and re-
sponse flow. It illustrates the manipulation of re-
sponses for plugin folder requests as described above.
3.2 Theme Detection
Theme detection is identical to plugin detection, only
with a change of the folder (’/wp-content/themes/’).
The version is detected through the style.css’ file and
the version tag.
Besides finding all installed themes, WPScan
looks for the ’main theme’. The ’main theme’ is the
currently active theme that styles the page. To de-
tect this theme, WPScan scans a website’s content for
links containing the themes folder.
To evade theme detection, we used the same two
methods as for the plugins, as described in Section
3.1. This includes logging fake themes.
Evading the main theme detection would require
changing links inside the website, which would lead
to a loss of the theme style. Since this is a significant
loss of functionality, we decided to not change it.
3.3 Version Detection
Version detection refers to detecting the running ver-
sion of the WordPress installation. WP displays the
version in different ways. WPScan can detect the fol-
lowing:
WordPress Head: WordPress prints the version in
the head of the website.
WordPress Generator: WordPress adds the ver-
sion the the XML generator.
Version Query: WordPress also shows the version
as a query argument for core style and script files.
Fingerprinting: All core JavaScript, CSS and
JSON files of a WordPress version have a unique
hash with which the version can be detected.
File Content: Both the ’install.php’ and
’load styles.php’ scripts output the version.
To break the version detection, we removed the WP
head and generator as well as the version query argu-
ments. We added a space to all the core files to change
the hashes, so fingerprinting is no longer possible. Fi-
nally, we removed the lines of code in the scripts, thus
they no longer display the versions.
Regarding simulation, it should not be possible
to manipulate the WordPress head or version query.
In theory, it is possible to change the generator re-
sponse. Fingerprinting could be simulated by adding
core files from older versions. This is, however, not
desirable since WordPress should be up to date. The
php file content of the ’install.php’ could be changed
instead of removed, but the ’load styles.php’ appends
versions to the links to get the correct file and there-
fore also could not be simulated.
3.4 User Enumeration
User enumeration tries to detect existing usernames
to log into the WordPress instance. For this WPScan
uses different functionality that WordPress offers:
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
488
request /wp-content/plugins/XXXPlugin Folder 'XXX' exists?
404 Response
403
Response
No
Yes
404
Plugin Enumeration
Fake Plugin
+ Logging
Hide Plugin
Wrong Plugin Results
403
WordPress Instance WPScan
SCANTRAP
Figure 1: Demonstration of response manipulation of request for plugin folder ’XXX’ with WP plugin ’SCANTRAP’.
REST API: WordPress allows to query users
through the available REST API.
JSON API: The JSON API also makes it possible
to query users.
Author Class: When authors of posts answer
comments, they are marked with an author class.
RSS Author Tag: The RSS feed contains a specific
tag naming the author.
Login Error: Login error messages can display if
the user exists.
URL Queries: WordPress allows to do URL
queries with the user id.
Author URL: WordPress adds an author link to
posts that contains the username.
Completly evading user enumeration comes with a
loss of functionality. The REST and JSON API are
disabled, which restricts the ability of applications to
interact with your WordPress site. URL queries are
forbidden. The author class, author link and RSS au-
thor tag are removed, which reduces the visibility of
authors. All login errors are disabled, which might
make troubleshooting more difficult. This could po-
tentially also effect search engine optimization.
With the exception of the RSS Author Tag, all
other areas can not be used for simulation purpose
since it would restrict the functionality. It might be
possible to create a user purposely as kind of a ’hon-
eypot’ user. However, the plugin removes all user
enumeration. The RSS Author Tag though could be
overwritten with an arbitrary name.
4 DISCUSSION
We tested our plugin in a test environment.
Details and the WPScan output can be found
in the Git repository (https://github.com/dfki-in-
sec/SCANTRAP). The results show that the proposed
strategies work for all four scans, with the exception
of the detection of the main theme, as described in
Section 3.2.
One downside regarding our methodology is that
some of the functionality, that might be desired by
the users, are restricted with the use of our plugin.
Especially user enumeration leads to potentially un-
desirable loss. However, user enumeration could be
seen as more of a privacy concern. Every defender
would need to decide for themselves if they put more
importance on functionality or privacy. One excep-
tion to this assumption is that WordPress usernames
are also login names and WordPress does not restrict
login tries. This makes user enumeration interesting
in the context of login brute force attacks. As men-
tioned in 3.2, the main theme detection is one exam-
ple of full loss of the main functionality style’. How-
ever, it has to be said that themes only make up 6% of
the existing vulnerabilities (WPScan, 2022). A sim-
ilar problem could occur if plugins require URLs to
the plugin folder on any pages. Another problem that
could occur is with search engine optimization, since
the plugin removes author links and manipulates sta-
tus codes to certain folders and files.
We are aware of two limitations in this methodol-
ogy: the first is our testing environment, as it is not
a real WordPress blog. We tested with only a limited
amount of plugins, themes, users and content. Due to
a large number of existing plugins alone, it would not
be feasible to test everything. The second is the re-
strictions on CMSs and CMS scanners. Our plugin is
only focusing on WordPress and WPScan. As shown
in Section 2 there are more CMSs and scanners that
might require adjustment to our method.
Cyber deception is supposed to deter attackers and
make attacking harder. It will not remove vulnera-
bilities or definitely hinder an attacker from targeting
the CMS. Therefore, having other security measures,
like up-to-date software, is still required. However,
it might discourage attackers completely from see-
ing the CMS as a target. More likely, it will require
more resources for additional manual reconnaissance
or lead attackers to the wrong conclusion about exist-
SCANTRAP: Protecting Content Management Systems from Vulnerability Scanners with Cyber Deception and Obfuscation
489
ing vulnerabilities. This might be helpful as a defence
method for overlooked vulnerabilities.
5 RELATED WORK
While research in regards to securing web content and
CMSs has been conducted, we have found little in re-
gards to using cyber deception, specifically for CMSs.
Jagamogan et al. (Jagamogan et al., 2021) con-
ducted a review on different penetration testing ap-
proaches on CMS. It underlines the likelihood of
CMSs being targeted. While the focus lies on ex-
ploitation frameworks, it is apparent that a lot are
based on different scanning methods. Specifically, it
is mentioned that WPScan and other CMS scanners
were used for attacks.
The paper by Jerkovi
´
c et al. (Jerkovi
´
c et al., 2016)
describes in more detail the problems and different
vulnerabilities of CMSs. Such as easy CMS identifi-
cation and a large base of unsecured and out-of-date
plugins. For protection, the paper lists different best
practices, such as content delivery networks (CDN),
regular updating and logging. They conclude that an
up-to-date CMS is not sufficient enough to prevent at-
tacks. Their solution for this is hosting in CDN, which
requires additional services and potential costs.
Efendi et al. (Efendi et al., 2019) highlighted the
use of deception techniques in the domain of web
applications. Their study describes web application
vulnerabilities and attacks and deception techniques.
Their focus lies on the detection of different attacks
on the application layer. The concentration is on ser-
vices and processes like SSH, Javascript and MySQL
and attacks such as Denial of Service and SQL Injec-
tions. In comparison, we are not focusing on detect-
ing specific attacks but rather hindrance of successful
reconnaissance and detection of such in the specific
domain of CMSs.
Valenza et al. (Valenza et al., 2020) developed a
prototype that attacks vulnerability scanners. In their
attacker model, the attacker becomes the victim and
the defender the attacker. The idea is to use vulnera-
bilities of the vulnerability scanner application to at-
tack it through HTTP responses. In regards to attacks,
their focus is on XSS and exploitation to get access to
the machine. Our paper is based on the same idea
that vulnerability scanners can not be attacked or ma-
nipulated. Instead of finding vulnerabilities in scan-
ners, however, our work focuses on finding structures
to manipulate the outcome of the scanner’s results.
A blog post from Medium Talk (Talk, 2022) de-
scribes two methods to hide the login URL. The first
is to change the name of the login file, the second is
to use a cookie to control access to the login page. A
similar approach to this is taken by the ’Hide My WP
Ghost - Security Plugin’ (Plugins, 2022). This plugin
allows changing and hiding all major paths and files
as well as some additional features.
The blog post from Cyberpunk (Cyberpunk, 2019)
that we have built our plugin on, is to our knowledge
one of a few comprehensive resources on using cy-
ber deception for CMSs. The post focuses specifically
and only on WordPress and WPScan. The first part of
the blog describes the usage of WPScan. The second
part goes into the defence against detection. For each
area, it gives a short description of how WPScan finds
information and a block of code that should prevent
it. Some of the defence methods require changes out-
side of WordPress. The focus of the blog post lies in
hiding information. We found the listings of sources
used by WPScan incomplete and therefore, some of
the code snippets did not lead to the desired effect
on their own. We build upon this by adding defence
against additional sources, removing the need to man-
ually change settings in web servers and adding upon
the idea by including additional cyber defence tech-
niques like simulation and honeytokens.
Another blog post that deals with WPScan obfus-
cation is by shift8 (Shift8, 2019). It deals specif-
ically with blocking WPScan with Nginx. It has
similar content and drawbacks to the previously de-
scribed blog post. However, it additionally mentions
the approach to blocking plugin and theme folders by
changing response codes to 404 status codes.
The first posts on blocking web CMS scanners, we
found, are about user enumeration (Andreasen, 2016)
and blocking WPScan (McRae, 2017). Both posts can
be considered partly outdated and incomplete. Both
focus on dissimulation and do not include simulation.
We have found a few plugins in the WordPress
store that have similar security and deception ideas.
Table 2 gives an overview of the different functions
offered by them. We are mainly looking at hiding and
faking, so dissimulation and simulation respectively.
Additionally, the detection and blocking of attackers
is a broad term for whitelisting, blacklisting IPs, as
well as preventing brute-force attacks. Some of the
plugins might also include some additional function-
ality.
’Blackhole for Bad Bots’ creates honeysites and
blocks IP addresses (Starr, 2022). In a similar manner,
’tinyShield’ (deprecated) adds functionality of white-
and blacklisting (tinyShield.me, 2022), so does ’No-
vaSense’ (NovaSense, 2020). Another plugin called
’BlogSafe Honeypot’ tracks failed URL requests and
login attempts (BlogSafe.org, 2021). ’Astra Security
Suite’ contains IP blocking mechanisms, logging, at-
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
490
Table 2: Comparing different WordPress security plugins regarding their dissimulation and simulation features.
Plugin / Capability
Hide/Fake
Users
Hide/Fake
Plugins
Hide/Fake
Themes
Hide/Fake
’wp-content’
Hide/Fake
Login Path
Hide/Fake
Login Error
Hide/Fake
Version
Hide/Fake
Cookies
Hide
WordPress
Hide/Fake
robots.txt
Detect/Block
Attackers
SCANTRAP /- / / -/- -/- /- /- -/- - -/- /-
Blackhole
(Starr, 2022) -/- -/- -/- -/- -/- -/- -/- -/- - -/ /
WP Ghost
(Plugins, 2022) /- /- /- /- / -/- -/- -/- -/- /
Don Security
(Donini, 2017) /- /- -/- -/- -/- -/- /- -/- - /- -/-
Stop User Enum
(Fullworks, 2022) /- -/- -/- -/- -/- -/- -/- -/- - -/- -/-
WP Smart Security
(WPSmartSecurity, 2015)
-/- /- /- /- /- /- / -/- -/- /
Titan
(CreativeMotion, 2022) -/- -/- -/- -/- /- -/- /- -/- - -/- /
tinyShield (deprecated)
(tinyShield.me, 2022) /- -/- -/- -/- -/- -/- -/- -/- - -/- /
block-wpscan
(rluisr, 2016) -/- -/- -/- -/- -/- -/- -/- -/- - -/- /
NovaSense
(NovaSense, 2020) -/- -/- -/- -/- -/- -/- -/- -/- - -/- /
Astra Security
(AstraSecurity, 2020) -/- -/- -/- -/- -/- -/- -/- -/- - -/- /
BlogSafe Honeypot
(BlogSafe.org, 2021) -/- -/- -/- -/- -/- -/- -/- -/- - -/- /-
tack prevention and more (AstraSecurity, 2020). This
is similar to the plugin ’WP Smart Security’, however,
it has not been updated in a while (WPSmartSecurity,
2015). The ’Titan Anti-spam & Security’ plugin does
spam detection, has a login rate limiter and uses WP-
Scan (CreativeMotion, 2022). The WPScan database
is also used by ’Jetpack Protect’ (Automatic, 2022).
Now, all of these plugins have their own technique
for securing WordPress with a strong focus on attack
prevention and less on scanner deception.
We found three plugins that focus more on scan-
ner evasion with similar methods to our plugin. ’Stop
User Enumeration’ focuses only on user enumera-
tion (Fullworks, 2022). The plugin ’block-wpscan’
is based on detecting browser languages and user
agents, however, it is no longer maintained and WP-
Scan can evade this type of detection (rluisr, 2016).
Finally, the plugin ’Don Security’ (Donini, 2017)
seems to be closest to our plugin. Its goal is to prevent
user enumeration, version detection and plugin detec-
tion, specifically for WPScan. It seems to be based on
some of the mentioned outdated blog posts and is also
no longer maintained.
6 CONCLUSIONS
In this paper, we have presented a WordPress plugin
that manipulates the output of WPScan scans, in or-
der to counter reconnaissance. This plugin can hide
the existence of plugins, themes and users from the
enumeration process and present false plugins and
themes. The results support the idea that cyber de-
ception and obfuscation methods are applicable to the
CMS enumeration process, and can also help to detect
attacks at an early stage. This could be effective to
protect from ’script kiddies’, bots or attackers that use
automation to gather information. Future work could
generalize this approach and transfer it to other CMS
than WordPress. It should be noted that scanning
tools could be adapted to be more resilient against cy-
ber deception, representing a cat-and-mouse game.
SCANTRAP: Protecting Content Management Systems from Vulnerability Scanners with Cyber Deception and Obfuscation
491
ACKNOWLEDGEMENTS
This research was supported by the German Federal
Ministry of Education and Research (BMBF) through
the Open6GHub project (Grant 16KISK003K).
REFERENCES
Andreasen, M. (2016). Protect + Stop WPScan
WordPress User Enumeration with Varnish.
https://guides.wp-bullet.com/protect-stop-wpscan-
wordpress-user-enumeration-varnish/. Accessed:
2022-09-28.
AstraSecurity (2020). Astra Security Suite Firewall
& Malware Scan. https://wordpress.org/plugins/
getastra/. Accessed: 2022-09-28.
Automatic (2022). Jetpack Protect. https://wordpress.org/
plugins/jetpack-protect/. Accessed: 2022-09-28.
Behera, C. K. and Bhaskari, D. L. (2015). Different Ob-
fuscation Techniques for Code Protection. Procedia
Computer Science, 70:757–763. Proceedings of the
4th International Conference on Eco-friendly Com-
puting and Communication Systems.
Bell, J. B. and Whaley, B. (1991). Cheating and Deception
(1st ed.). Transaction Publ, New Brunswick.
BlogSafe.org (2021). BlogSafe Honeypot. https://
wordpress.org/plugins/blogsafe-honeypot/. Accessed:
2022-09-28.
Boiko, B. (2005). Content Management Bible. Bible. Wiley.
CreativeMotion (2022). Titan Anti-spam & Security. https:
//wordpress.org/plugins/anti-spam/. Accessed: 2022-
09-28.
Cyberpunk (2019). WPScan Usage Example [Enumeration
+ Exploit]. https://www.cyberpunk.rs/wpscan-usage-
example. Accessed: 2022-09-21.
de Barros, A. (2007). Dlp and honeytokens.
Donini, R. (2017). Don Security. https://wordpress.org/
plugins/don-security/. Accessed: 2022-09-28.
Efendi, A. M., Ibrahim, Z., Zawawi, M. A., Abdul Rahim,
F., Pahri, N. M., and Ismail, A. (2019). A Survey on
Deception Techniques for Securing Web Application.
In 2019 IEEE 5th Intl Conference on Big Data Secu-
rity on Cloud, pages 328–331.
Fraunholz, D., Anton, S. D., Lipps, C., Reti, D., Krohmer,
D., Pohl, F., Tammen, M., and Schotten, H. D.
(2018a). Demystifying Deception Technology:A Sur-
vey.
Fraunholz, D., Reti, D., Duque Anton, S., and Schotten,
H. D. (2018b). Cloxy: A Context-Aware Deception-
as-a-Service Reverse Proxy for Web Services. In Pro-
ceedings of the 5th ACM Workshop on Moving Target
Defense, MTD ’18, page 40–47. ACM.
Fullworks (2022). Stop User Enumeration. https://
wordpress.org/plugins/stop-user-enumeration/. Ac-
cessed: 2022-09-28.
InfosecMatter (2020). CMS Vulnerability Scanners
for WordPress, Joomla, Drupal, Moodle, Typo3..
https://www.infosecmatter.com/cms-vulnerability-
scanners-for-wordpress-joomla-drupal-moodle-
typo3/. Accessed: 2022-09-21.
Jagamogan, R. S., Ismail, S. A., Hafizah, N., and
Hafiza Abas, H. (2021). A Review: Penetration
Testing Approaches on Content Management Sys-
tem (CMS). In 2021 7th International Conference
on Research and Innovation in Information Systems
(ICRIIS), pages 1–6.
Jerkovi
´
c, H., Vrane
ˇ
si
´
c, P., and Dadi
´
c, S. (2016). Secur-
ing web content and services in open source con-
tent management systems. In 2016 39th International
Convention on Information and Communication Tech-
nology, Electronics and Microelectronics (MIPRO),
pages 1402–1407.
McRae, S. (2017). Blocking the scanners WP-
Scan. https://gorgolak.wordpress.com/2017/05/27/
blocking-the-scanners-wpscan/. Accessed: 2022-09-
28.
NovaSense (2020). NovaSense. https://wordpress.org/
plugins/novasense-security/. Accessed: 2022-09-28.
Plugins, W. W. S. (2022). Hide My WP Ghost Se-
curity Plugin. https://de.wordpress.org/plugins/hide-
my-wp/. Accessed: 2022-10-05.
rluisr (2016). block-wpscan. https://wordpress.org/plugins/
block-wpscan/. Accessed: 2022-09-28.
Shift8 (2019). How to block your WordPress site
from being scanned by WPScan with Nginx.
https://shift8web.ca/2019/01/how-to-block-your-
wordpress-site-from-being-scanned-by-wpscan-
with-nginx/. Accessed: 2022-09-28.
Spitzner, L. (2003). The honeynet project: trapping the
hackers. IEEE Security & Privacy, 1(2):15–23.
Starr, J. (2022). Blackhole for Bad Bots. https:
//wordpress.org/plugins/blackhole-bad-bots/. Ac-
cessed: 2022-09-28.
Talk, M. (2022). How to Change The WordPress Login
URL Without Plugin. https://www.mediumtalk.net/
change-wordpress-login-url-without-plugin/. Ac-
cessed: 2022-10-05.
tinyShield.me (2022). tinyShield Simple. Focused. Se-
curity. https://wordpress.org/plugins/tinyshield/. Ac-
cessed: 2022-09-28 and 2022-12-19.
Valenza, A., Costa, G., and Armando, A. (2020). Never
Trust Your Victim: Weaponizing Vulnerabilities in Se-
curity Scanners.
W3Techs (2022). Usage statistics of content management
systems. https://w3techs.com/technologies/overview/
content\ management. Accessed: 2022-09-21.
WPScan (2022). WordPress Vulnerability Statistics. https:
//wpscan.com/statistics. Accessed: 2022-09-20.
WPScanTeam (2022). WPScanTeam WPScan Github.
https://github.com/wpscanteam/wpscan. Accessed:
2022-09-21.
WPSmartSecurity (2015). WP Smart Security. https://
wordpress.org/plugins/wp-smart-security/. Accessed:
2022-09-28.
Yuill, J. J. et al. (2007). Defensive Computer-Security
Deception Operations: Processes, Principles and
Techniques. https://repository.lib.ncsu.edu/handle/
1840.16/5648. Accessed: 2022-09-21.
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
492