/ctf-forensics

CTF Forensics - Disk and Memory Analysis

Table of Contents

• Memory Forensics (Volatility 3)
• Disk Image Analysis
• VM Forensics (OVA/VMDK)
• VMware Snapshot Forensics
• GIMP Raw Memory Dump Visual Inspection (INShAck 2018)
• Coredump Analysis
• Windows KAPE Triage Analysis (UTCTF 2026)
• PowerShell Ransomware Analysis
• Android Forensics
• Container Forensics (Docker)
• Cloud Storage Forensics (AWS S3 / GCP / Azure)
• BSON (Binary JSON) Format Reconstruction (IceCTF 2016)
• TrueCrypt / VeraCrypt Volume Mounting (GreHack CTF 2016)
• Volatility mftparser Offset-Based Deleted File Recovery (BSides Delhi 2018)
• Brotli Blob Detection via ASCII-Art Signature (ASIS Finals 2018)
• corkami/pocs MD5 PDF Collision Generation (35C3 2018)
• See Also

Memory Forensics (Volatility 3)

vol -f memory.dmp windows.info
vol -f memory.dmp windows.pslist
vol -f memory.dmp windows.cmdline
vol -f memory.dmp windows.netscan
vol -f memory.dmp windows.filescan
vol -f memory.dmp windows.dumpfiles --physaddr <addr>
vol -f memory.dmp windows.mftscan | grep flag

Common plugins:

• windows.pslist / windows.pstree - Process listing
• windows.cmdline - Command line arguments
• windows.netscan - Network connections
• windows.filescan - File objects in memory
• windows.dumpfiles - Extract files by physical address
• windows.mftscan - MFT FILE objects in memory (timestamps, filenames). Note: mftparser was Volatility 2 only; Vol3 uses mftscan

Disk Image Analysis

# Mount read-only
sudo mount -o loop,ro image.dd /mnt/evidence

# Autopsy / Sleuth Kit
fls -r image.dd              # List files recursively
icat image.dd <inode>        # Extract file by inode

# Carving deleted files
photorec image.dd
foremost -i image.dd

VM Forensics (OVA/VMDK)

# OVA = TAR archive containing VMDK + OVF
tar -xvf machine.ova

# 7z reads VMDK directly (no mount needed)
7z l disk.vmdk | head -100
7z x disk.vmdk -oextracted "Windows/System32/config/SAM" -r

Key files to extract from VM images:

• Windows/System32/config/SAM - Password hashes
• Windows/System32/config/SYSTEM - Boot key
• Windows/System32/config/SOFTWARE - Installed software
• Users/*/NTUSER.DAT - User registry
• Users/*/AppData/ - Browser data, credentials

VMware Snapshot Forensics

Converting VMware snapshots to memory dumps:

# .vmss (suspended state) + .vmem (memory) → memory.dmp
vmss2core -W path/to/snapshot.vmss path/to/snapshot.vmem
# Output: memory.dmp (analyzable with Volatility/MemprocFS)

Malware hunting in snapshots (Armorless):

1. Check Amcache for executed binaries near encryption timestamp
2. Look for deceptive names (Unicode lookalikes: ṙ instead of r)
3. Dump suspicious executables from memory
4. If PyInstaller-packed: pyinstxtractor → decompile .pyc
5. If PyArmor-protected: use PyArmor-Unpacker

Ransomware key recovery via MFT:

• Even if original files deleted, MFT preserves modification timestamps
• Seed-based encryption: recover mtime → derive key

vol -f memory.dmp windows.mftscan | grep flag
# mtime as Unix epoch → seed for PRNG → derive encryption key

GIMP Raw Memory Dump Visual Inspection (INShAck 2018)

Pattern: When Volatility fails or profiles don't match, open raw memory dumps directly in GIMP as raw image data. Scroll through memory while adjusting image width to find previously-displayed images rendered as pixel data.

Steps:

1. Open .dmp file in GIMP: File > Open, set image type to "Raw image data"
2. Set pixel format to RGB, width to ~1920 (monitor resolution)
3. Scroll through the file offset while adjusting width with arrow keys
4. Previously-displayed images (desktop, browser content) become visible when width matches the original stride

# Alternative: use Python + PIL to scan memory as pixel data
python3 -c "
from PIL import Image
import numpy as np

with open('memory.dmp', 'rb') as f:
    data = f.read()

# Try common display widths: 1920, 1366, 1280, 1024
for width in [1920, 1366, 1280, 1024]:
    stride = width * 3  # RGB = 3 bytes per pixel
    # Sample at various offsets through the dump
    for offset in range(0, len(data) - stride * 100, stride * 500):
        chunk = data[offset:offset + stride * 100]
        if len(chunk) == stride * 100:
            img = Image.frombytes('RGB', (width, 100), chunk)
            # Check if image has meaningful content (not all zeros/noise)
            arr = np.array(img)
            if 10 < arr.mean() < 245 and arr.std() > 20:
                img.save(f'frame_{width}_{offset}.png')
                print(f'Potential image at offset {offset}, width {width}')
"

Key insight: Raw memory dumps contain framebuffer data that was displayed on screen. GIMP can render arbitrary binary data as pixels. When the image width matches the original display stride, screenshots of the user's desktop become visible without any forensics tools, profiles, or decryption.

Coredump Analysis

gdb -c core.dump
(gdb) info registers
(gdb) x/100x $rsp
(gdb) find 0x0, 0xffffffff, "flag"

Windows KAPE Triage Analysis (UTCTF 2026)

Pattern (Landfall, Sherlockk, Cold Workspace): KAPE (Kroll Artifact Parser and Extractor) triage collection ZIP containing Windows forensic artifacts. Multiple challenges reference the same triage dataset.

KAPE triage structure:

Modified_KAPE_Triage_Files/
├── C/
│   ├── Users/<username>/
│   │   ├── AppData/Local/Microsoft/Windows/PowerShell/PSReadLine/
│   │   │   └── ConsoleHost_history.txt    # PowerShell command history
│   │   ├── NTUSER.DAT                     # User registry hive
│   │   └── AppData/Roaming/Microsoft/Windows/Recent/  # Recent files
│   ├── Windows/
│   │   ├── System32/config/
│   │   │   ├── SAM          # Password hashes
│   │   │   ├── SYSTEM       # System config + boot key
│   │   │   └── SOFTWARE     # Installed software
│   │   └── appcompat/Programs/
│   │       └── Amcache.hve  # Execution history with SHA-1 hashes
│   └── $MFT                 # Master File Table
└── ...

High-value artifacts:

1. PowerShell history — reveals attacker commands:

cat "C/Users/*/AppData/Local/Microsoft/Windows/PowerShell/PSReadLine/ConsoleHost_history.txt"
# Look for: credential access, lateral movement, data staging

2. Amcache — executed programs with timestamps and hashes:

# Parse with Eric Zimmerman's AmcacheParser or regipy
python3 -c "
from regipy.registry import RegistryHive
reg = RegistryHive('C/Windows/appcompat/Programs/Amcache.hve')
for entry in reg.recurse_subkeys(as_json=True):
    print(entry)
" | grep -i "flag\|suspicious\|malware"

3. MFT resident data — small files stored directly in MFT records:

# Parse MFT for resident file data (files < ~700 bytes stored inline)
# Use analyzeMFT or python-ntfs
import struct

with open('$MFT', 'rb') as f:
    mft_data = f.read()

# Search for flag patterns in raw MFT data
import re
flags = re.findall(rb'utflag\{[^}]+\}', mft_data)
for flag in flags:
    print(f"Found: {flag.decode()}")

4. Environment variables from memory dumps (Cold Workspace pattern):

# Small .dmp files may be minidumps with environment variable blocks
strings -a cold-workspace.dmp | grep -i "flag\|password\|key\|secret"
# Environment variables survive in process memory snapshots

Challenge patterns from UTCTF 2026:

• Landfall: Flag hidden in PowerShell history or Amcache execution records
• Sherlockk: Correlate Amcache entries with MFT timestamps to identify malicious activity
• Cold Workspace: Flag in environment variables extracted from memory dump
• Checkpoint A/B: Multi-part investigation using combined artifacts

Key insight: KAPE triage ZIPs contain pre-collected forensic artifacts — no need for full disk imaging. Start with PowerShell history (fastest wins) → Amcache (execution timeline) → MFT (resident data for small files) → registry hives (persistence, credentials).

PowerShell Ransomware Analysis

Pattern (Email From Krampus): PowerShell memory dump + network capture.

Analysis workflow:

1. Extract script blocks from minidump:

python power_dump.py powershell.DMP
# Or: strings powershell.DMP | grep -A5 "function\|Invoke-"

2. Identify encryption (typically AES-CBC with SHA-256 key derivation)

3. Extract encrypted attachment from PCAP:

# Filter SMTP traffic in Wireshark
# Export attachment, base64 decode

4. Find encryption key in memory dump:

# Key often generated with Get-Random, regex search:
strings powershell.DMP | grep -E '^[A-Za-z0-9]{24}$' | sort | head

5. Find archive password similarly, decrypt layers

Android Forensics

# Extract APK from device
adb pull /data/app/com.target.app/base.apk

# Analyze APK contents
apktool d base.apk -o decompiled/
# Check: AndroidManifest.xml, res/values/strings.xml, shared_prefs/

# Extract data from Android backup
adb backup -apk -shared -all -f backup.ab
java -jar abe.jar unpack backup.ab backup.tar
tar xf backup.tar

# SQLite databases (contacts, messages, browser history)
sqlite3 /data/data/com.android.providers.contacts/databases/contacts2.db ".tables"
sqlite3 /data/data/com.android.providers.telephony/databases/mmssms.db "SELECT * FROM sms"

# Parse Android filesystem image
mkdir android_mount && mount -o ro android_image.img android_mount/
# Key locations:
# /data/data/<app>/databases/     — app SQLite databases
# /data/data/<app>/shared_prefs/  — app preferences (XML)
# /data/system/packages.xml       — installed packages
# /data/misc/wifi/wpa_supplicant.conf — saved WiFi passwords

Key insight: Android stores app data in /data/data/<package>/ with SQLite databases and XML shared preferences. adb backup captures the full app state. For CTFs, check shared_prefs/ for hardcoded secrets and databases/ for flags.

Container Forensics (Docker)

# Export Docker image layers
docker save IMAGE:TAG -o image.tar
tar xf image.tar
# Each layer is a directory with layer.tar containing filesystem changes
# Check: layer.tar files for added/modified files, deleted files (.wh.* whiteout)

# Inspect image history for build commands (may contain secrets)
docker history IMAGE:TAG --no-trunc
# Shows every Dockerfile instruction including ARGs and ENV values

# Extract filesystem without running the container
docker create --name extract IMAGE:TAG
docker export extract -o container_fs.tar
docker rm extract

# Analyze with dive (layer-by-layer diff viewer)
dive IMAGE:TAG

# Common forensic targets in container images:
# /app/.env, /app/config/* — application secrets
# /root/.bash_history     — build-time commands
# /etc/shadow             — leaked credentials
# Deleted files visible in earlier layers even if removed in later ones

Key insight: Docker images are layered — a file deleted in a later layer still exists in the earlier layer's tar. Use docker history --no-trunc to see full Dockerfile commands including secrets passed via ARG or ENV. The dive tool visualizes layer diffs interactively.

Cloud Storage Forensics (AWS S3 / GCP / Azure)

# Enumerate public S3 buckets
aws s3 ls s3://target-bucket/ --no-sign-request
aws s3 cp s3://target-bucket/flag.txt . --no-sign-request

# Check bucket versioning (previous versions may contain deleted flags)
aws s3api list-object-versions --bucket target-bucket --no-sign-request
aws s3api get-object --bucket target-bucket --key secret.txt --version-id VERSION_ID out.txt

# GCP Cloud Storage
gsutil ls gs://target-bucket/
gsutil cp gs://target-bucket/flag.txt .

# Azure Blob Storage
az storage blob list --container-name target --account-name storageaccount
az storage blob download --container-name target --name flag.txt --account-name storageaccount

Key insight: Cloud storage versioning preserves deleted objects. Even if a flag file is deleted from the bucket, previous versions may still be accessible via list-object-versions. Always check for versioning-enabled buckets.

BSON (Binary JSON) Format Reconstruction (IceCTF 2016)

BSON is MongoDB's binary serialization format. Corrupted BSON files need header repair before parsing, and may contain base64-encoded file fragments.

import bson

# BSON header: first 4 bytes = little-endian document size
# Fix corrupted header by setting correct size
with open('data.bson', 'rb') as f:
    data = bytearray(f.read())

# Fix size header if corrupted (e.g., missing first 3 bytes)
import struct
correct_size = len(data) + 3  # account for missing bytes
data = struct.pack('<I', correct_size)[1:] + data  # prepend missing bytes

# Parse BSON documents
docs = bson.decode_all(bytes(data))
for doc in docs:
    print(doc)

# Reconstruct file from BSON chunks (common pattern):
# Each document has: {index: N, data: "base64_chunk"}
import base64
chunks = sorted(docs, key=lambda d: d.get('index', d.get('i', 0)))
reconstructed = b''
for chunk in chunks:
    b64_data = chunk.get('data', chunk.get('d', ''))
    reconstructed += base64.b64decode(b64_data)

with open('reconstructed.png', 'wb') as f:
    f.write(reconstructed)

Key insight: BSON starts with a 4-byte little-endian size field. If the file appears corrupted, check if the first bytes are missing or incorrect. Parse with bson.decode_all() (from pymongo), sort chunks by index, and concatenate base64-decoded data to reconstruct embedded files.

TrueCrypt / VeraCrypt Volume Mounting (GreHack CTF 2016)

Encrypted volumes in CTF challenges may use TrueCrypt or VeraCrypt. Identify by logo/branding clues, then mount with a recovered keyfile or password.

# Identify TrueCrypt volumes:
# - No file signature/magic bytes (by design)
# - Exact size is multiple of 512 bytes
# - High entropy throughout the file
# - Context clues: TrueCrypt logo in related images

# Mount with password:
truecrypt -t -p "password123" volume.tc /mnt/tc
veracrypt -t -p "password123" volume.tc /mnt/vc

# Mount with keyfile (no password):
truecrypt -t -p "" -k keyfile.png volume.tc /mnt/tc
veracrypt -t -p "" -k keyfile.png volume.tc /mnt/vc

# Mount hidden volume (different password):
truecrypt -t -p "hidden_password" volume.tc /mnt/tc

# Common keyfile locations in CTFs:
# - Images extracted from other challenge steps
# - GPG-encrypted files with keys found in git repos
# - Files embedded in other forensic artifacts

# If TrueCrypt is not available (discontinued):
# Use VeraCrypt (backwards-compatible with TrueCrypt volumes)
# Add --truecrypt flag for old TC volumes:
veracrypt -t --truecrypt -p "password" volume.tc /mnt/vc

Key insight: TrueCrypt volumes have no magic bytes or identifiable header -- they look like random data. Identify them from context clues (related images showing TrueCrypt logo, file sizes that are exact multiples of 512, or challenge descriptions mentioning encryption). VeraCrypt with --truecrypt flag handles legacy TC volumes.

Volatility mftparser Offset-Based Deleted File Recovery (BSides Delhi 2018)

Pattern: Standard dumpfiles or filescan + dumpfiles --physaddr fails on a deleted file because its directory entry has been marked free. The MFT record that still holds the file's $DATA attribute survives until the record is reused. Volatility 2's mftparser can dump the resident $DATA directly when given the exact --offset of the MFT record found via filescan.

# 1. Locate the MFT record offset (Volatility 2 example; Vol3 uses windows.mftscan)
vol.py -f Challenge.raw --profile=Win7SP1x86 mftparser \
    | grep -A2 "target_filename"

# 2. Dump every attribute of the matching MFT entry, including $DATA
vol.py -f Challenge.raw --profile=Win7SP1x86 mftparser \
    --offset=0x7ca3c00 --dump-dir=./out/

ls ./out/
# file.data.$DATA contains the recovered content

Key insight: NTFS marks files "deleted" by flipping one bit in the MFT record header (0x16 byte: 0x01 == in use). Until the record is reallocated, the entire $DATA attribute is still intact and only lazily freed. Use mftparser --offset=<record> for resident files (under ~700 bytes — stored inline in the MFT) and dd/icat with the cluster runs for larger files. Always also grep for the filename in windows.mftscan output before giving up: memory-resident MFT fragments are still findable after on-disk deletion.

References: BSides Delhi CTF 2018 — Never Too Late Mister, writeups 11963, 11970

Brotli Blob Detection via ASCII-Art Signature (ASIS Finals 2018)

Pattern: Binwalk and file miss Brotli-compressed data because the format has no fixed magic. Decompress candidate blobs with brotli.decompress(); the Brotli reference implementation embeds its own ASCII-art logo Brrroootttllliii as a sanity-check output. If the decompressed bytes contain that or other Brotli-specific telemetry strings, the original blob was Brotli.

import brotli
try:
    out = brotli.decompress(blob)
    if b'rrrooottl' in out or b'Brotli' in out:
        print('Brotli-compressed')
except Exception: pass

Key insight: Any compressor without a magic byte is identifiable by trial decompression. For Brotli, zstd, snappy, lzma-alone, spin through each library in order until one succeeds without raising.

References: ASIS CTF Finals 2018 — Green Cabbage, writeup 12419

corkami/pocs MD5 PDF Collision Generation (35C3 2018)

Pattern: Challenge demands two valid PDFs with the same MD5 but different content. Use corkami/pocs pdf.py combined with enscript | ps2pdf to produce a PDF header with collision-friendly padding, then drive the collision via fastcoll (or hashclash for chosen-prefix). Works because the PDF format tolerates garbage in the %PDF trailer region that the MD5 collision block can overwrite.

enscript -p out.ps content.txt
ps2pdf out.ps base.pdf
python pdf.py base.pdf target1.pdf target2.pdf
fastcoll -p base.pdf -o target1.pdf target2.pdf
md5sum target1.pdf target2.pdf    # identical

Key insight: PDF collision is a one-command pipeline with the right toolchain. The harder variant is chosen-prefix MD5 (different visible content) which requires hashclash and 10-20 CPU-hours. Check pocs/collisions/ for every file format with prebuilt scaffolds.

References: 35C3 CTF 2018 — collider, writeup 12836

See Also

• disk-advanced.md - Advanced disk and memory techniques (deleted partition recovery, ZFS forensics, GPT GUID encoding, VMDK sparse parsing, memory dump string carving, ransomware key recovery, WordPerfect macro XOR, minidump ISO 9660 recovery, APFS snapshot recovery, RAID 5 XOR recovery, Kyoto Cabinet hash DB forensics)
• disk-recovery.md - Disk recovery and extraction patterns (LUKS master key recovery, PRNG timestamp seed brute-force, VBA macro binary recovery, FemtoZip decompression, XFS reconstruction, tar duplicate entry extraction, nested matryoshka filesystem extraction, anti-carving via null byte interleaving)