Skip to main content

Flawed CSRF token implementation

The sole purpose of (secret) CRSF token is to help the application identify authenticated or unauthenticated requests. Any request that doesn’t contain csrf tokens are treated as unauthenticated one thus rejected by the application as the csrf tokens are only available to the authenticated users.
But contrary to that, in one application, the csrf tokens are generated before login and worse, it’s not regenerated after successful authentication of the user. This defeats the purpose of anti-csrf approach.

Anti-CSRF best practices:

    Don’t issue csrf tokens before authentication
    Always regenerate the tokens after successful authentication, if issued before authentication
    Use POST methods for critical transactions embedding csrf tokens
    Don’t send the csrf tokens in GET requests as they may reveal it in browser logs etc

In few Ruby based apps, where the token was being generated before authentication and same was being used.



Mitigation: (taken from 'Symbolic Security Analysis of Ruby-on-Rails Web Applications' white paper):

While talking about csrf protection, they say, Finally, we must account for “insider attacks,” i.e., attacks by users of the application (against other users of the application). To understand this issue, we need to look again at the implementation of token generation on line 30 above. The complication here is that session[:_csrf_token] is not reset automatically by Rails between logins, hence different users that log in from the same IP address could inadvertently be given the same CSRF token. To properly protect against CSRF, the application should always change session[:_csrf_token] to nil before logging in a user, so that the token is regenerated whenever a different user logs in. Rails provides a method, reset_session, which has just this effect:

class Controller::Base # continued

    def reset_session() session = {} end

end

Comments

Popular posts from this blog

Ardilla- New tool for finding SQL Injection and XSS

Three Researchers -- MIT's Adam Kiezun , Stanford's Philip Guo , and Syracuse University's Karthick Jayaraman -- has developed a new tool ' Ardilla ' that automatically finds and exploits SQL injection and cross-site scripting vulnerabilities in Web applications. It creates inputs that pinpoint bugs in Web applications and then generates SQL injection and XSS attacks. But for now Ardilla is for PHP -based Web app only. The researchers say Ardilla found 68 never-before found vulnerabilities in five different PHP applications using the tool -- 23 SQL injection and 45 XSS flaws. More information is awaited. For their attack generation techniques refer to their document at: http://www.cs.washington.edu/homes/mernst/pubs/create-attacks-tr054.pdf

File Upload through Null Byte Injection

Sometimes, during file upload we come across situation wherein there would be check on the file extension at the client side as well as server side too. If the application does allow only .jpeg extension to be uploaded, the client side java script checks for the extension of the file before passing the request. We all know that how easily this can be defeated. Some applications, checks for the extension at the server side also. That's not easy to bypass. However there are some ways with which it still can be bypassed. Most of server side scripts are written in high level languages such as Php, Java etc who still use some C/C++ libraries to read the file name and contents. That leads to the problem. In C/C++ a line ends with /00 or which is called Null Byte. So whenever the interpreter sees a null byte at the end of the a string, it stops reading thinking it has reached at the end of the string. This can be used for the bypass. It works for many servers, specially php servers. T

Combining power of Fiddler with Burp

Both are pretty powerful tools when it comes to intercept and modify http communications. But at some point of time, they become even more powerful combo if tied with each other. They complement each other. In a recent pentest I came across a similar situation where in Burp was not able to intercept a specific kind of traffic and Fiddler came to rescue. The application was designed to upload video. The initial communication was straight forward, I mean logging into application, filling up the video details etc. And all these were easily captured by Burp except the point where you hit the Upload Video and it connects to a different server and surprisingly it was not captured by Burp, not sure why, even after repeated attempts. So, I fired Fiddler to see if the it sees this request. But it's a;ways to play with requests using Burp due to it's various functionalities like, Intruder, Repeaters etc. But it was necessary to capture this request in Burp. So the below steps can be