Cookies Hacking
Last updated
Last updated
Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)
Cookies come with several attributes that control their behavior in the user's browser. Here’s a rundown of these attributes in a more passive voice:
The expiry date of a cookie is determined by the Expires attribute. Conversely, the Max-age attribute defines the time in seconds until a cookie is deleted. Opt for Max-age as it reflects more modern practices.
The hosts to receive a cookie are specified by the Domain attribute. By default, this is set to the host that issued the cookie, not including its subdomains. However, when the Domain attribute is explicitly set, it encompasses subdomains as well. This makes the specification of the Domain attribute a less restrictive option, useful for scenarios where cookie sharing across subdomains is necessary. For instance, setting Domain=mozilla.org makes cookies accessible on its subdomains like developer.mozilla.org.
A specific URL path that must be present in the requested URL for the Cookie header to be sent is indicated by the Path attribute. This attribute considers the / character as a directory separator, allowing for matches in subdirectories as well.
When two cookies bear the same name, the one chosen for sending is based on:
The cookie matching the longest path in the requested URL.
The most recently set cookie if the paths are identical.
The SameSite attribute dictates whether cookies are sent on requests originating from third-party domains. It offers three settings:
Strict: Restricts the cookie from being sent on third-party requests.
Lax: Allows the cookie to be sent with GET requests initiated by third-party websites.
None: Permits the cookie to be sent from any third-party domain.
Remember, while configuring cookies, understanding these attributes can help ensure they behave as expected across different scenarios.
Request Type
Example Code
Cookies Sent When
Link
<a href="..."></a>
NotSet*, Lax, None
Prerender
<link rel="prerender" href=".."/>
NotSet*, Lax, None
Form GET
<form method="GET" action="...">
NotSet*, Lax, None
Form POST
<form method="POST" action="...">
NotSet*, None
iframe
<iframe src="..."></iframe>
NotSet*, None
AJAX
$.get("...")
NotSet*, None
Image
<img src="...">
NetSet*, None
Table from Invicti and slightly modified. A cookie with SameSite attribute will mitigate CSRF attacks where a logged session is needed.
*Notice that from Chrome80 (feb/2019) the default behaviour of a cookie without a cookie samesite attribute will be lax (https://www.troyhunt.com/promiscuous-cookies-and-their-impending-death-via-the-samesite-policy/). Notice that temporary, after applying this change, the cookies without a SameSite policy in Chrome will be treated as None during the first 2 minutes and then as Lax for top-level cross-site POST request.
This avoids the client to access the cookie (Via Javascript for example: document.cookie)
If the page is sending the cookies as the response of a requests (for example in a PHPinfo page), it's possible to abuse the XSS to send a request to this page and steal the cookies from the response (check an example in https://hackcommander.github.io/posts/2022/11/12/bypass-httponly-via-php-info-page/.
This could be Bypassed with TRACE HTTP requests as the response from the server (if this HTTP method is available) will reflect the cookies sent. This technique is called Cross-Site Tracking.
This technique is avoided by modern browsers by not permitting sending a TRACE request from JS. However, some bypasses to this have been found in specific software like sending \r\nTRACE instead of TRACE to IE6.0 SP2.
Another way is the exploitation of zero/day vulnerabilities of the browsers.
It's possible to overwrite HttpOnly cookies by performing a Cookie Jar overflow attack:
It's possible to use Cookie Smuggling attack to exfiltrate these cookies
The request will only send the cookie in an HTTP request only if the request is transmitted over a secure channel (typically HTTPS).
Cookies prefixed with __Secure- are required to be set alongside the secure flag from pages that are secured by HTTPS.
For cookies prefixed with __Host-, several conditions must be met:
They must be set with the secure flag.
They must originate from a page secured by HTTPS.
They are forbidden from specifying a domain, preventing their transmission to subdomains.
The path for these cookies must be set to /.
It is important to note that cookies prefixed with __Host- are not allowed to be sent to superdomains or subdomains. This restriction aids in isolating application cookies. Thus, employing the __Host- prefix for all application cookies can be considered a good practice for enhancing security and isolation.
So, one of the protection of __Host- prefixed cookies is to prevent them from being overwritten from subdomains. Preventing for example Cookie Tossing attacks. In the talk Cookie Crumbles: Unveiling Web Session Integrity Vulnerabilities (paper) it's presented that it was possible to set __HOST- prefixed cookies from subdomain, by tricking the parser, for example, adding "=" at the beggining or at the beginig and the end...:
Or in PHP it was possible to add other characters at the beginning of the cookie name that were going to be replaced by underscore characters, allowing to overwrite __HOST- cookies:
If a custom cookie contains sensitive data check it (specially if you are playing a CTF), as it might be vulnerable.
Sensitive data embedded in cookies should always be scrutinized. Cookies encoded in Base64 or similar formats can often be decoded. This vulnerability allows attackers to alter the cookie's content and impersonate other users by encoding their modified data back into the cookie.
This attack involves stealing a user's cookie to gain unauthorized access to their account within an application. By using the stolen cookie, an attacker can impersonate the legitimate user.
In this scenario, an attacker tricks a victim into using a specific cookie to log in. If the application does not assign a new cookie upon login, the attacker, possessing the original cookie, can impersonate the victim. This technique relies on the victim logging in with a cookie supplied by the attacker.
If you found an XSS in a subdomain or you control a subdomain, read:
Cookie TossingHere, the attacker convinces the victim to use the attacker's session cookie. The victim, believing they are logged into their own account, will inadvertently perform actions in the context of the attacker's account.
If you found an XSS in a subdomain or you control a subdomain, read:
Cookie TossingClick on the previous link to access a page explaining possible flaws in JWT.
JSON Web Tokens (JWT) used in cookies can also present vulnerabilities. For in-depth information on potential flaws and how to exploit them, accessing the linked document on hacking JWT is recommended.
This attack forces a logged-in user to execute unwanted actions on a web application in which they're currently authenticated. Attackers can exploit cookies that are automatically sent with every request to the vulnerable site.
(Check further details in theoriginal research) Browsers permit the creation of cookies without a name, which can be demonstrated through JavaScript as follows:
The result in the sent cookie header is a=v1; test value; b=v2;. Intriguingly, this allows for the manipulation of cookies if an empty name cookie is set, potentially controlling other cookies by setting the empty cookie to a specific value:
This leads to the browser sending a cookie header interpreted by every web server as a cookie named a with a value b.
In Chrome, if a Unicode surrogate codepoint is part of a set cookie, document.cookie becomes corrupted, returning an empty string subsequently:
This results in document.cookie outputting an empty string, indicating permanent corruption.
(Check further details in theoriginal research) Several web servers, including those from Java (Jetty, TomCat, Undertow) and Python (Zope, cherrypy, web.py, aiohttp, bottle, webob), mishandle cookie strings due to outdated RFC2965 support. They read a double-quoted cookie value as a single value even if it includes semicolons, which should normally separate key-value pairs:
(Check further details in theoriginal research) The incorrect parsing of cookies by servers, notably Undertow, Zope, and those using Python's http.cookie.SimpleCookie and http.cookie.BaseCookie, creates opportunities for cookie injection attacks. These servers fail to properly delimit the start of new cookies, allowing attackers to spoof cookies:
Undertow expects a new cookie immediately after a quoted value without a semicolon.
Zope looks for a comma to start parsing the next cookie.
Python's cookie classes start parsing on a space character.
This vulnerability is particularly dangerous in web applications relying on cookie-based CSRF protection, as it allows attackers to inject spoofed CSRF-token cookies, potentially bypassing security measures. The issue is exacerbated by Python's handling of duplicate cookie names, where the last occurrence overrides earlier ones. It also raises concerns for __Secure- and __Host- cookies in insecure contexts and could lead to authorization bypasses when cookies are passed to back-end servers susceptible to spoofing.
According to this blogpost, it might be possible to use the cookie attribute $Version=1 to make the backend use an old logic to parse the cookie due to the RFC2109. Moreover, other values just as $Domain and $Path can be used to modify the behaviour of the backend with the cookie.
This parsing indicate to unescape escaped values inside the cookies, so "\a" becomes "a". This can be useful to bypass WAFS as:
eval('test') => forbidden
"\e\v\a\l\(\'\t\e\s\t\'\)" => allowed
In the RFC2109 it's indicated that a comma can be used as a separator between cookie values. And also it's possible to add spaces and tabs before an after the equal sign. Therefore a cookie like $Version=1; foo=bar, abc = qux doesn't generate the cookie "foo":"bar, admin = qux" but the cookies foo":"bar" and "admin":"qux". Notice how 2 cookies are generated and how admin got removed the space before and after the equal sign.
Finally different backdoors would join in a string different cookies passed in different cookie headers like in:
Which could allow to bypass a WAF like in this example:
The cookie is the same every time you login.
Log out and try to use the same cookie.
Try to log in with 2 devices (or browsers) to the same account using the same cookie.
Check if the cookie has any information in it and try to modify it
Try to create several accounts with almost the same username and check if you can see similarities.
Check the "remember me" option if it exists to see how it works. If it exists and could be vulnerable, always use the cookie of remember me without any other cookie.
Check if the previous cookie works even after you change the password.
If the cookie remains the same (or almost) when you log in, this probably means that the cookie is related to some field of your account (probably the username). Then you can:
Try to create a lot of accounts with usernames very similar and try to guess how the algorithm is working.
Try to bruteforce the username. If the cookie saves only as an authentication method for your username, then you can create an account with username "Bmin" and bruteforce every single bit of your cookie because one of the cookies that you will try will the one belonging to "admin".
Try Padding Oracle (you can decrypt the content of the cookie). Use padbuster.
Padding Oracle - Padbuster examples
Padbuster will make several attempts and will ask you which condition is the error condition (the one that is not valid).
Then it will start decrypting the cookie (it may take several minutes)
If the attack has been successfully performed, then you could try to encrypt a string of your choice. For example, if you would want to encrypt user=administrator
This execution will give you the cookie correctly encrypted and encoded with the string user=administrator inside.
CBC-MAC
Maybe a cookie could have some value and could be signed using CBC. Then, the integrity of the value is the signature created by using CBC with the same value. As it is recommended to use as IV a null vector, this type of integrity checking could be vulnerable.
The attack
Get the signature of username administ = t
Get the signature of username rator\x00\x00\x00 XOR t = t'
Set in the cookie the value administrator+t' (t' will be a valid signature of (rator\x00\x00\x00 XOR t) XOR t = rator\x00\x00\x00
ECB
If the cookie is encrypted using ECB it could be vulnerable. When you log in the cookie that you receive has to be always the same.
How to detect and attack:
Create 2 users with almost the same data (username, password, email, etc.) and try to discover some pattern inside the given cookie
Create a user called for example "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa" and check if there is any pattern in the cookie (as ECB encrypts with the same key every block, the same encrypted bytes could appear if the username is encrypted).
There should be a pattern (with the size of a used block). So, knowing how are a bunch of "a" encrypted you can create a username: "a"*(size of the block)+"admin". Then, you could delete the encrypted pattern of a block of "a" from the cookie. And you will have the cookie of the username "admin".
Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)
