SignSaboteur is a Burp Suite extension for editing, signing, verifying, and attacking signed tokens. It supports different types of tokens: Django TimestampSigner, ItsDangerous Signer, Express cookie-session middleware, OAuth2 Proxy, Tornado’s signed cookies, Ruby Rails Signed cookies, Ruby Rails Encrypted cookies, Nimbus JOSE + JWT and Unknown signed string.
You can find more information about the extension on the Portswigger Research blog post page - Introducing SignSaboteur: forge signed web tokens with ease.
It provides automatic detection and in-line editing of token within HTTP requests/responses and WebSocket messages, signing of tokens and automation of brute force attacks against signed tokens implementations.
It was inspired by Fraser Winterborn and Dolph Flynn JWT Token extension. The original source code can be found here and here.
- Ensure that Java JDK 17 or newer is installed
- From root of project, run the command
./gradlew jar
- This should place the JAR file
sign-saboteur-1.0.6.jar
within thebuild/libs
directory - This can be loaded into Burp by navigating to the
Extensions
tab,Installed
sub-tab, clickingAdd
and loading the JAR file - This BApp is using the newer Montoya API, so it's best to use the latest version of Burp (try the earlier adopter channel if there are issues with the latest stable release)
The Wordlist View
allows to import secrets and salts list files. Extension has own prebuild dictionary lists. Most
secrets are taken from jwt-secrets. As an
option, Flask-Unsign-Wordlist can be used. Extension supports JSON
strings format for special chars, to use it quot the secret string with "
.
The Editor View
supports a number of signed tokens: Django, Dangerous, Flask, Express, OAuth2 and Tornado. It allows
modification of the signed tokens at Burp Suite's HTTP Request/Response view in the Proxy, History and Repeater tools.
The Dangerous tab can be used for both, Flask
and Django
tokens, which are selected depending on whether a Dangerous
or Django token is detected.
The Unknown tab can be used to brute force unknown signed strings. Guessing mode works only with Balanced and Deep brute force attacks. It supports different message derivation technics, including:
- None message will be used as is
- CONCAT separator byte will be removed from the message and that new value will be used to calculate signature
- Tornado separator byte will be added to the end of the message string
A JSON text editor is provided to edit each component that contain JSON content:
- Dangerous Payload
- Django Payload (except pickle serialized payload)
- Express Payload
- JWT Payload
A timestamp editor is provided to edit each component that contain it:
- Dangerous timestamp
- Django timestamp
- OAuth2 Proxy timestamp
- Tornado timestamp
A hex editor is provided to all signed tokens, except Express signatures. NOTE Express Tab doesn't support signature auto update yet. Please copy it manually to corresponding signature cookie.
Sign
presents a signing dialog that can be used to update the Signature by signing the token using a key from
the Keys View
that has signing capabilities
Brute force
will attempt to find secret key that was used for signature generation. If a secret key was found, a
dialog will be presented.
The Brute force
option implements three types of attacks against signed tokens Signatures:
- Known keys will use previously found secret keys only
- Fast will use default hashing algorithm and key derivation
- Balanced will use all known key derivation technics, except PBKDF2WithHmacSHA1, PBKDF2WithHmacSHA256
- Deep will use all key derivation technics, including different types supported by Ruby Rails framework
The Attack
option implements eight well-known authorization attacks against signed tokens:
- User claims
- Wrapped user claims
- Username and password claims
- Flask claims
- Express claims
- Account user claims
- Authenticated claims
- User access_token
These are described in more detail below.
All of these attacks can be used together. Please bear in mind that extension doesn't support token payload modification at Attack mod, so your payload will be replaced with new one.
OpenID connect ID token format usually used by signing libraries to store information about authenticated user.
Extension will generate placeholder for admin
user ID token.
Same as User claims
attack but will put it into user
JSON attribute.
Another common way to store user details is username
and password
JSON attributes. Extension will generate
placeholder for admin
user.
Flask authenticated user session information if stored at client side should include id
and _id
and user_id
or _user_id
JSON attribute. Extension will generate session for the first user, usually admin.
Express framework uses passport
JSON attribute to store user details. Extension will generate placeholder for admin
user.
Some frameworks may use account
wrapper to store information about authenticated user. For exploitation, it might be
required to use Authenticated claims
too.
The Authenticated claims implements 12 well-known authorization flags.
The User access_token
option generated JWT OpenID connect ID token signed with the same key and same hashing algorithm
without any key derivation preformed.