This exploit was developed during pentesting activity against oracle OAM 11.1.2.3.0. It was developed based on the technical description by sec-consult
https://www.sec-consult.com/en/blog/2018/05/oracle-access-managers-identity-crisis/
The exploit depend on python-paddingoracle it can be downloaded from here
https://github.com/mwielgoszewski/python-paddingoracle
# python oracle-oam-exploit.py -h
####### # # # #######
# # # # ## ## # # # ##### # #### # #####
# # # # # # # # # # # # # # # # # #
# # # # # # # ##### ## # # # # # # #
# # ####### # # # ## ##### # # # # #
# # # # # # # # # # # # # # #
####### # # # # ####### # # # ###### #### # #
Oracle Padding Oracle
coded by: Mostafa Soliman
usage: oracle-oam-exploit.py [-h] [-e ENCRYPT] [-d DECRYPT] [-v] URL
positional arguments:
URL Target resource URL
optional arguments:
-h, --help show this help message and exit
-e ENCRYPT, --encrypt ENCRYPT
Encrypt plain text data
-d DECRYPT, --decrypt DECRYPT
Decrypt base64 encode cipher text
-v, --verb Show decrypt block info
The exploit recieve the URL of the resource which require oracle OAM authentication. It does the below steps:
The pentester can specify any encrypted value (encquery, encreplay, cookie) he wish to decrypt using the -d flag.
The pentester can specify any plain text value he wish to encrypt using the -e flag.
by reversing oracle OAM sdk we can see that the cookie consists of the below structure
Salt= ACL= AuthId= Ip= TCT= SessionId= userId= validate=BASE64Encode(MD5(Salt+AuthId+Ip+ACL+TCT+SessionId))
So to build a cookie, get a valid plain text one using -d flag, modify it then encrypt it again using -e.
