Hi guys, this is the first writeup about my vulnerability bounty program,a process about how I discovered a Twitter XSS vulnerability.
I think that in the process of finding the vulnerability, there are some interesting knowledge points, I hope you can get some from my writeup.
If you want to know more details, you need to visit bobrov blog, my discovery is due to reading his writeup, and thanks bobrov very much,I have a lot of gains from his blog.
Maybe you don’t want to spend more time. Here I will give a brief explanation of his article. When you visit some addresses, the server returns 302, which is similar to the following picture.
In the returned Body, location will choose how to populate according to the requested URL,and the requested URI will be placed in the href event.
What do you think of next? Can we try it with
indeed, as you might expect,
So, bobrov did a lot of fuzz, trying to get around the limitations here, amazing, he did, his Fuzz process is as follows (taken from his blog)
So, his final Payload is
and the response from the server is as follows.Location header =
//aa :1/:/dev.twitter.com/%01xxx . returned body =
http:////aa:1/:/dev.twitter.com/%01xxx,does the browser not jump?
There is a key point in this.When you visit a website, usually, the port you access is 80 or 8080,if you enter the port address as 1, the browser will not jump.Here, it is the a tag. The click was made possible, so we can combine the issue via Clickjacking,let the victim trigger this XSS vulnerability
After introducing his findings, let me talk about how I found this problem again.Honestly, I have been paying attention to Twitter security for a while,but I have gained very little,so I put my eye on Twitter’s past vulnerability report to see if there are some problems in the Twitter fix.
When I saw bobrov writeup about the problem,I immediately visited dev.twtter.com to try to find out if there were some problems.When I collected information through Google, I found an extremely interesting link.
Look closely, do you find it interesting? My feeling tells me that there may be problems here.So I tried to go to Fuzz to confirm my guess.After a period of testing, I found that when I added a / at the end of the URL,a miracle happened, for example,
when I visited
Yes,Twitter’s fix for this issue no longer exists. Finally,after a period of Fuzz, my final Payload was
Twitter 1120$ rewards
My gains in the process:
Don’t give up on the exploitation of a large company’s loopholes, insist that there will always be gains
When accessing a 302 jump page, location header and link on page extracting Url in different ways, there may be problems here.
When the browser tries to access a website with a port of 1, it will not jump
I hope that you can make some gains from my writing and sorry I m a non English speaker