178 lines
No EOL
7.4 KiB
Markdown
178 lines
No EOL
7.4 KiB
Markdown
SNIPs of SIGINT :: (U) Monthly Notes for June 2012
|
|
|
|
DYNAMIC PAGE
|
|
System Max: SECRET//SI/TALENT KEYHOLE//REL TO USA, FVEY
|
|
Portion Max: TOP SECRET//SI//REL TO USA, FVEY
|
|
|
|
SNIPS of SIGINT
|
|
(U) Monthly Notes for June 2012
|
|
|
|
1 of 5
|
|
|
|
SNIPS of SIGINT :: (U) Monthly Notes for June 2012
|
|
TOP SECRET//SI//REL TO USA, FVEY
|
|
(U) Monthly Notes for June 2012
|
|
|
|
2012.06.27 - 03:58 pm
|
|
Entry tags: monthly notes
|
|
|
|
Technical accomplishments
|
|
|
|
(U//FOUO) PITIED FOOL
|
|
|
|
* (TS//SI//REL) PITIEDFOOL is a suite of CNA tools I'm developing for
|
|
use against file systems, initially focused on Windows. More information
|
|
can be found on the DSD wiki, but it's a bit ugly at this point.
|
|
* (TS//SI//REL) Determined that it was a combination of the volume
|
|
shadow copy and deleted entries in the shadow copy that were causing
|
|
my recovery woes. As a result of the shadow copy having large chunks of
|
|
MFT data in it, and much of these records still being mostly correct,
|
|
recovery programs were able to reconstruct most of the system drive. I
|
|
added functionality to PITIEDFOOL to overwrite the contents of the
|
|
shadow copy on each partition encountered.
|
|
* (TS//SI//REL) Now that I was overwriting large chunks of data for
|
|
every partition, I could no longer ignore the performance issues with
|
|
writing to a disk one sector at a time. I refactored my code to overwrite in
|
|
blocks configurable by a pre-processor definition and sped up execution
|
|
about 15-fold, with the possiblity of it going even faster with increased
|
|
memory usage.
|
|
* (TS//SI//REL) After all this I tested PITIEDFOOL within FUZZYEBOLA
|
|
and it worked flawlessly. I took a build of FUZZYEBOLA from last
|
|
month, and without recompiling inserted the PITIEDFOOL binary with
|
|
configuration details to execute it at a certain time. At that time I saw the
|
|
process usage slightly increase (from 0% to around 2%) and a few minutes
|
|
later the system rebooted and didn't come back up. Running a file
|
|
recovery tool over the entire drive yielded some files (from scraping
|
|
headers) but nearly the entire contents of the drive were irrecoverable,
|
|
and if it had been configured to securely wipe every sector on the drive
|
|
after killing the MFT and VSS it wouldn't have been able to recover
|
|
anything at all. Success!
|
|
* (U//FOUO) Documented my work.
|
|
|
|
(U//FOUO) PANT SPARTY
|
|
|
|
* (TS//SI//REL) PANT SPARTY is a backdoor in the SSH daemon for
|
|
\*NIX, based on OpenSSH portable. It allows a public key to be embedded
|
|
in the sshd binary and will then always grant a root login shell if
|
|
presented with the proper key pair for that key. In other words, it behaves
|
|
as if the given key is in ~/.ssh/authorized_keys.
|
|
* (TS//SI//REL) Currently DSD uses authorized_keys as a quick-and-easy
|
|
method of persistence against certain \*NIX targets. In most cases this
|
|
|
|
2 of 5
|
|
|
|
SNIPS of SIGINT :: (U) Monthly Notes for June 2012
|
|
|
|
works, but it obviously isn't very stealthy and can run into problems if
|
|
that file is deleted or the SSH configuration changes to not use it. The goal
|
|
for this project was to provide the same level of persistence but embedded
|
|
in the sshd binary itself (obviously, assuming root access, as before).
|
|
* (TS//SI//REL) By adding my own check in several key places within the
|
|
SSH code I was able to grant authorization to a key embedded in the
|
|
binary without too much trouble. The hard part came in granting that
|
|
user root privileges even if the configuration specifically banned root
|
|
login. That was achieved by setting a flag in the authorization context
|
|
within the server that essentially said "this is our backdoor, let it happen,
|
|
oh and stop logging also."
|
|
* (TS//SI//REL) Then I wanted to have it so if you provide a normal
|
|
username as a login, it still grants you a root shell. SSH has a lot of checks
|
|
to make sure you can't switch usernames in the middle of a login (go
|
|
figure) so this was a bit tricky to bypass. Ultimately I was able to use the
|
|
same flag as the previous step, and just had to put checks for it in other
|
|
places.
|
|
* (TS//SI//REL) More fun was to be had when I wanted to allow an
|
|
arbitrary username to be provided (i.e. one that doesn't exist on the
|
|
system) while still allowing for root login. This led to all sorts of problems
|
|
where I didn't even get a valid authorization context at all, and I couldn't
|
|
manually call the C function to get one for root because the connection is
|
|
a de-privileged child process, so I had to force-feed it a fake one for root.
|
|
Of course, the force-fed hash for root doesn't match the entry in passwd,
|
|
so I had to throw in another check there to return true if it's the backdoor.
|
|
And there's all sorts of fun happening with pre-auth where I had to revert
|
|
the authorization context back to its original value to postpone its
|
|
processing (a standard authorization step in some cases) and then
|
|
properly update it to root during the actual authorization step later on.
|
|
Once I got this working, however, a user is able to run "ssh
|
|
abcdefg@target" and the target's logs only show "Invalid username
|
|
abcdefg" without the follow-on message of "Accepted publickey for root
|
|
from 10.0.0.1 port 55000 ssh2\". Presenting a valid username leads to the
|
|
only logged output being that the public key check failed.
|
|
* (S//SI//REL) I set my new implant development PR on this project! My
|
|
previous record was 3 days for DIRTYDEEDS, but I finished this in 2.
|
|
Technically I had it working in a day, but it wasn't what I considered
|
|
releasable, so I'll have to wait for some other opportunity to present itself
|
|
so I can achieve the feat of an implant-in-a-day.
|
|
|
|
Awards/Recognition
|
|
|
|
(U//FOUO) Letter of Appreciation from [REDACTED] Director DSD
|
|
|
|
3 of 5
|
|
|
|
SNIPS of SIGINT :: (U) Monthly Notes for June 2012
|
|
|
|
(U//FOUO) Letter of Appreciation from [REDACTED] Chief SUSLOC
|
|
|
|
(U//FOUO) SUSLOC coin
|
|
|
|
Briefings and demonstrations given
|
|
|
|
(S//REL) Briefed the U.S. Ambassador to Australia on my work on CNA/CNE
|
|
and collaboration with TAO during my time at DSD.
|
|
|
|
Meetings and briefings attended
|
|
|
|
(U) NSTR
|
|
|
|
(U) Weekly SUSLOC meetings
|
|
|
|
Reports/papers/publications
|
|
|
|
(U) NSTR
|
|
|
|
Training
|
|
|
|
(U) NSTR
|
|
|
|
Miscellaneous
|
|
|
|
(U) New Zealand was incredible! I wish I'd had more time there, but I think I
|
|
did pretty well. I saw a handful of LOTR sights, Mount Cook, a number of
|
|
gorgeous lakes, snow-capped mountains everywhere, [REDACTED]
|
|
[REDACTED] If you want pictures let me know and I'll shoot
|
|
you the link.
|
|
|
|
(U) Tasmania was similarly awesome, and not just because I can now say "Yea,
|
|
I've been to Tasmania," because, you know, it sounds cool.
|
|
|
|
(U) redacted at the U.S. Embassy Independence Day Party!
|
|
|
|
(U//FOUO) My time at DSD came to a close in June, as my flight back is on 30
|
|
June (26 hours in an airplane, WOO!). I absolutely loved my time in Australia,
|
|
both in terms of work and travel, but I'm also looking forward to returning to
|
|
the land of Chick-fil-A, college athletics, BBQ pork, and real bacon. Oh, and
|
|
good beer.
|
|
|
|
(U) THE BCS IS DEAD!!!!!!! YESSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
|
|
SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
|
|
|
|
4 of 5
|
|
|
|
SNIPS of SIGINT :: (U) Monthly Notes for June 2012
|
|
|
|
SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
|
|
SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS!!!
|
|
|
|
(Leave a comment)
|
|
|
|
[REDACTED]
|
|
Page Publisher: T314 - End User Solutions
|
|
|
|
DERIVED FROM: NSA/CSSM 1-52
|
|
DATED: 08 JAN 2007
|
|
DECLASSIFY ON: 20320108
|
|
DYNAMIC PAGE
|
|
System Max: SECRET//SI/TALENT KEYHOLE//REL TO USA, FVEY
|
|
Portion Max: TOP SECRET//SI//REL TO USA, FVEY
|
|
|
|
5 of 5 |