3v1lw1th1n/ircollect

ircollect

ircollect

ircollect is a Python tool designed to collect files of interest in an incident response investigation or triage effort. Running as local admin, it:

• Opens the raw disk
• Reads the master boot record, collects a copy of it, and uses the MBR to find partition and disk information
• Using the MBR information, it finds the NTFS partitions.
• Working from the start of the NTFS partition, it finds the $MFT
• It collects a copy of the $MFT and then builds a list of all the files on the system and their data runs
• Using the file list and data runs, it collects interesting files through direct reads from the disk, bypassing access controls.

All collected files are stored in a directory specified with the -d option. They are further organized by hostname and the date-time the script was run.

Requirements:

pip install analyzemft

Status:

VERY beta. Active development daily, often hourly.

Currently collects master boot record, $MFT, and live (corrupted) registry hives. User can modify table in ircollect.py to specify any files they desire.

Upcoming features:

File an issue, please, if there is something you'd like.

• Windows executable
• Netcat support
• Handle Win7 vs XP hive location differences
• Add configuration file support
• Collect uncorrupted registry hives

Notes:

Unless I am missing something, you cannot recover the live registry hives through direct disk reads. Well, you can, but they are corrupted. I'm currently getting them via a created snapshot. But, if you're going to create a snapshot, you might as well get all of the interesting files from the snapshot.... Yes? No? Could rebuild the corrupted hives, but that is a completely different problem.

Bugs:

Many

Thank you to:

• Jamie Levy for mbr_parser
• Willi Ballenthin - bit manipulation code, lots of useful tips for analyzeMFT

Questions and Answers:

Q: Will this read any file, regardless of access controls? A: Yes.

Q: Doesn't this tool give malicious actors access to information they shouldn't have? A: This isn't rocket science. The method used is documented in many places and several other tools use it. Not publishing this out of fear of misuse is doing a disservice to those who can really use it to help stop malicious actors.