Awesome
Understanding CSRF
The Express team's csrf and csurf modules frequently have issues popping up concerned about our usage of cryptographic functions. These concerns are unwarranted due to a misunderstanding of how CSRF tokens work. So here's a quick run down!
Read this and still have questions? Want to tell us we're wrong? Open an issue!
How does a CSRF attack work?
On their own (phishing site), an attacker could create an AJAX button or form that creates a request against your site:
<form action="https://my.site.com/me/something-destructive" method="POST">
<button type="submit">Click here for free money!</button>
</form>
This is worse with AJAX as the attacker could use other methods like DELETE
as well as read the result.
This is particularly important when the user has some sort of session with very personal details on your site.
If this is in the context of a technologically illiterate user,
there might be some inputs for credit card and social security info.
How to mitigate CSRF attacks?
Use only JSON APIs
AJAX calls use JavaScript and are CORS-restricted.
There is no way for a simple <form>
to send JSON
,
so by accepting only JSON,
you eliminate the possibility of the above form.
Disable CORS
The first way to mitigate CSRF attacks is to disable cross-origin requests.
If you're going to allow CORS,
only allow it on OPTIONS, HEAD, GET
as they are not supposed to have side-effects.
Unfortunately, this does not block the above request as it does not use JavaScript (so CORS is not applicable).
Check the referrer header
Unfortunately, checking the referrer header is a pain in the ass, but you could always block requests whose referrer headers are not from your site. This really isn't worth the trouble.
For example, you could not load sessions if the referrer header is not your server.
GET should not have side effects
Make sure that none of your GET
requests change any relevant data in your database.
This is a very novice mistake to make and makes your app susceptible to more than just CSRF attacks.
Avoid using POST
Because <form>
s can only GET
and POST
,
by using other methods like PUT
, PATCH
, and DELETE
,
an attacker has fewer methods to attack your site.
Don't use method override!
Many applications use method-override to use
PUT
, PATCH
, and DELETE
requests over a regular form.
This, however, converts requests that were previously invulnerable vulnerable!
Don't use method-override
in your apps - just use AJAX!
Don't support old browsers
Old browsers do not support CORS or security policies. By disabling support for older browsers (which more technologically-illiterate people use, who are more (easily) attacked), you minimize CSRF attack vectors.
CSRF Tokens
Alas, the final solution is using CSRF tokens. How do CSRF tokens work?
- Server sends the client a token.
- Client submits a form with the token.
- The server rejects the request if the token is invalid.
An attacker would have to somehow get the CSRF token from your site, and they would have to use JavaScript to do so. Thus, if your site does not support CORS, then there's no way for the attacker to get the CSRF token, eliminating the threat.
Make sure CSRF tokens can not be accessed with AJAX!
Don't create a /csrf
route just to grab a token,
and especially don't support CORS on that route!
The token just needs to be "unguessable", making it difficult for an attacker to successfully guess within a couple of tries. It does not have to be cryptographically secure. An attack is one or two clicks by an unbeknownst user, not a brute force attack by a server.
BREACH attack
This is where the salt comes along.
The BREACH attack is pretty simple: if the server sends the same or very similar response over HTTPS+gzip
multiple times,
an attacker could guess the contents of response body (making HTTPS utterly useless).
Solution? Make each response a tiny bit different.
Thus, CSRF tokens are generated on a per-request basis and different every time. But the server needs to know that any token included with a request is valid. Thus:
- Cryptographically secure CSRF tokens are now the CSRF "secret", (supposedly) only known by the server.
- CSRF tokens are now a hash of the secret and a salt.
Read more here:
Note that CSRF doesn't solve the BREACH attack, but the module simply randomizes requests to mitigate the BREACH attack for you.
The salt doesn't have to be cryptographically secure
Because the client knows the salt!!!
The server will send <salt>;<token>
and the client will return the same value to the server on a request.
The server will then check to make sure <secret>+<salt>=<token>
.
The salt must be sent with the token,
otherwise the server can't verify the authenticity of the token.
This is the simplest cryptographic method. There are more methods, but they are more complex and not worth the trouble.
Creating tokens must be fast
Because they are created on every request!
Doing something as simple as Math.random().toString(36).slice(2)
is sufficient as well as extremely performant!
You don't need OpenSSL to create cryptographically secure tokens on every request.
The secret doesn't have to be secret
But it is.
If you're using a database-backed session store,
the client will never know the secret as it's stored on your DB.
If you're using cookie sessions,
the secret will be stored as a cookie and sent to the client.
Thus, make sure cookie sessions use httpOnly
so the client can't read the secret via client-side JavaScript!
When you're using CSRF tokens incorrectly
Adding them to JSON AJAX calls
As noted above, if you don't support CORS and your APIs are strictly JSON, there is absolutely no point in adding CSRF tokens to your AJAX calls.
Exposing your CSRF token via AJAX
Don't ever create a GET /csrf
route in your app
and especially don't enable CORS on it.
Don't send CSRF tokens with API response bodies.
Conclusion
As the web moves towards JSON APIs and browsers become more secure with more security policies, CSRF is becoming less of a concern. Block older browsers from accessing your site and change as many of your APIs to be JSON APIs, and you basically no longer need CSRF tokens. But to be safe, you should still enable them whenever possible and especially when it's non-trivial to implement.