Sniffly is an attack that abuses HTTP Strict Transport Security and Content Security Policy to allow arbitrary websites to sniff a user's browsing history. It has been tested in Firefox and Chrome.
More info available in the ToorCon 2015 slides: https://zyan.scripts.mit.edu/presentations/toorcon2015.pdf .
Demo
Visit http://zyan.scripts.mit.edu/sniffly/ in Firefox/Chrome/Opera with HTTPS Everywhere disabled. If you use an ad blocker, a bunch of advertising domains will probably show up in the "Probably Visited" column (ignore them).
How it works
I recommend reading the inline comments in src/index.js to understand how Sniffly does a timing attack in both FireFox and Chrome without polluting the local HSTS store.
- User visits Sniffly page
- Browser attempts to load images from various HSTS domains over HTTP
- Sniffly sets a CSP policy that restricts images to HTTP, so image sources are blocked before they are redirected to HTTPS. This is crucial! If the browser completes a request to the HTTPS site, then it will receive the HSTS pin, and the attack will no longer work when the user visits Sniffly.
- When an image gets blocked by CSP, its onerror handler is called. In this case, the onerror handler does some fancy tricks to time how long it took for the image to be redirected from HTTP to HTTPS. If this time is on the order of a millisecond, it was an HSTS redirect (no network request was made), which means the user has visited the image's domain before. If it's on the order of 100 milliseconds, then a network request probably occurred, meaning that the user hasn't visited the image's domain.
Finding HSTS hosts
To scrape an included list of sites ( util/strict-transport-security.txt , courtesy Scott Helme) to determine which hosts send HSTS headers, do:
$ cd util
$ ./run.sh <number_of_batches> > results.log
where 1 batch is 100 sites. You can override util/strict-transport-security.txt with a different list, such as the full Alexa Top 1M, if you want.
To process and sort the results by max-age, excluding ones with max-age less than 1 day and ones that are preloaded:
$ cd util
$ ./process.py <results_file> > processed.log
Once that's done, you can copy the hosts from processed.log into src/index.js .
Running sploitz
Visiting file:///path/to/sniffly/src/index.html in Chrome should just work. In Firefox, CSP headers using the tag are apparently not supported yet, so you need to set up a local webserver to serve the CSP HTTP response header. My Nginx server block looks something like this:
server {
listen 8081;
server_name localhost;
location / {
root /path/to/sniffly/src;
add_header Content-Security-Policy "img-src http:";
index index.html;
}
}
Or in .htaccess :
<IfModule mod_headers.c>
Header set Content-Security-Policy "img-src http:"
</IfModule>
Or send the header via php .
Paste this at the start of the script (and change the name to index.php):
<?php
$csp_rules = "img-src http:";
// Just to ensure maximum compatibility
header('X-WebKit-CSP: '.$csp_rules);
header('X-Content-Security-Policy: '.$csp_rules);
header('Content-Security-Policy: '.$csp_rules);
?>
Caveats
- Not supported yet in Safari, IE, or Chrome on iOS.
- Extensions such as HTTPS Everywhere will mess up results.
- Doesn't work reliably in Tor Browser since timings are rounded to the nearest 100-millisecond.
- Users with a different HSTS preload list (ex: due to having an older browser) may not see accurate results.
Acknowledgements
- Scott Helme for an initial list of HSTS hosts that he had found so I didn't have to scan the entire Alexa 1M.
- Chris Palmer for advising on how to file a privacy bug in Chrome.
- Dan Kaminsky and WhiteOps for sponsoring the ToorCon trip where this was presented.
- Jan Schaumann and Chris Rohlf for being early testers.
- Everyone who let me sleep on their couch while I did this over my "vacation break". You know who you are!
This posting includes an audio/video/photo media file: Download Now
No comments:
Post a Comment