Skip to content
Back to skills

Ftk

ASecurity

Automate authorized FTK evidence workflows with FTK Imager, distributed processing, and known-file hash filtering.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 22, 2026
ai-agentsgoshellbashsqlnodegitdatabasebackendsecurity

Works with

  • cursor
  • cli

Security analysis

A100/100

Pro scans all 3 files and shows the line behind each finding

Scanned October 22, 2026

npx -y skills add alivirgo/Major-AI-Skills --skill ftk --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ftk?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Ftk
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/alivirgo-ftk/badge)](https://www.skillsdirectory.com/skills/alivirgo-ftk)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: ftk
description: "Automate authorized FTK evidence workflows with FTK Imager, distributed processing, and known-file hash filtering."
category: digital-forensics
risk: safe
source: self
source_type: self
date_added: "2026-08-26"
tags: ["ftk", "ftk-imager", "forensic-toolkit", "dpe", "kff-filter", "digital-forensics", "claude"]
tools: ["claude", "cursor", "gemini", "codex"]
---

# Exterro FTK (Forensic Toolkit) AI Skill Guide (Claude)

## Overview & Engine Architecture
Exterro FTK (Forensic Toolkit) is an enterprise-grade digital forensics platform known for its **Distributed Processing Engine (DPE)**, centralized PostgreSQL database backend, and **FTK Imager** forensic acquisition suite. FTK processes multi-terabyte evidence sets simultaneously across worker clusters, integrating the **Known File Filter (KFF)**, Cerberus malware triage, and automated OCR. Claude operates as a Senior Forensic Systems Engineer and DFIR Architect, specializing in **FTK Imager CLI automation**, **DPE cluster troubleshooting**, **volatile memory acquisition (RAM)**, and **case database optimization**.

### FTK Distributed Processing & Database Architecture

```
┌─────────────────────────────────────────────────────────────┐
│                 FTK Enterprise Architecture                 │
│                                                             │
│  Central Core & Database Layer                              │
│  ├── PostgreSQL Database Cluster (Case Meta & Artifact DB)  │
│  ├── KFF (Known File Filter - NSRL Hash Set Library)        │
│  └── Cerberus Malware Analysis & Scoring Engine             │
│                                                             │
│  Distributed Processing & Acquisition Layer                 │
│  ├── Distributed Processing Engine (DPE Worker Nodes - 5000)│
│  ├── FTK Imager CLI (`ftkimager.exe` E01/RAW Bitstream Acq)│
│  └── Optical Character Recognition (OCR Engine Subsystem)   │
└─────────────────────────────────────────────────────────────┘
```

---

## Operational Capabilities & Agent Directives

1. **FTK Imager CLI Automation**: Script automated forensic acquisitions of physical disks (`\\.\PhysicalDrive0`) and volatile physical memory into E01 containers with embedded cryptographic verification (MD5/SHA-1).
2. **Distributed Processing Engine (DPE) Optimization**: Configure and troubleshoot DPE worker node communication over TCP port 5000, adjusting thread worker allocations and PostgreSQL connection pools.
3. **KFF & Hash Library Management**: Maintain Known File Filter (KFF) hash libraries, importing NIST NSRL databases to automatically eliminate benign operating system binaries from investigator review.
4. **Live RAM & Volatile Triage**: Safely acquire live RAM dumps from Windows workstations, diagnosing kernel driver blocking caused by Windows Virtualization-Based Security (VBS/HVCI).

---

## Production PowerShell Automation: Automated FTK Imager Live Acquisition

Save this script as `Run-FTKAcquisition.ps1` to automate physical drive imaging and volatile RAM acquisition via FTK Imager CLI:

```powershell
<#
.SYNOPSIS
    Automated Forensic Live Acquisition using FTK Imager CLI (ftkimager.exe)
.DESCRIPTION
    Dumps volatile physical RAM and acquires PhysicalDrive0 into E01 container with verification.
#>
param (
    [Parameter(Mandatory=$true)]
    [string]$CaseName,

    [Parameter(Mandatory=$true)]
    [string]$DestinationDir
)

$FTK_CLI = "C:\Program Files\AccessData\FTK Imager\ftkimager.exe"

if (-not (Test-Path $FTK_CLI)) {
    Write-Error "FTK Imager CLI not found at '$FTK_CLI'."
    exit 1
}

New-Item -ItemType Directory -Force -Path $DestinationDir | Out-Null
$Timestamp = Get-Date -Format "yyyyMMdd_HHmmss"

# 1. Acquire Volatile Physical Memory (RAM)
$RamOut = Join-Path $DestinationDir "${CaseName}_${Timestamp}_RAM.raw"
Write-Host "[1/2] Acquiring Physical Memory (RAM) to: $RamOut..." -ForegroundColor Cyan
& $FTK_CLI --print-physical-memory $RamOut

# 2. Acquire Physical Drive 0 to E01 Container
$DiskOut = Join-Path $DestinationDir "${CaseName}_${Timestamp}_Disk"
Write-Host "[2/2] Acquiring PhysicalDrive0 to E01 format: $DiskOut.E01..." -ForegroundColor Cyan
& $FTK_CLI "\\.\PhysicalDrive0" $DiskOut `
    --e01 `
    --compress 6 `
    --frag 2048M `
    --case-number "$CaseName" `
    --evidence-number "001" `
    --examiner "DFIR Lead" `
    --description "Automated Live Triage Disk Image" `
    --verify

if ($LASTEXITCODE -eq 0) {
    Write-Host "Forensic acquisition completed and cryptographically verified!" -ForegroundColor Green
} else {
    Write-Error "Acquisition failed with exit code $LASTEXITCODE"
}
```

---

## Technical Troubleshooting Matrix

| Issue & Failure Signature | Root Cause Analysis | Diagnostic & Resolution Pathway |
| :--- | :--- | :--- |
| **DPE Worker Node Disconnects / Fails Registration** | Windows Firewall blocked TCP port 5000 or the worker service lost PostgreSQL database credentials. | 1. In `services.msc`, verify **AccessData Distributed Processing Engine** service is Running.<br>2. Open inbound TCP port **5000** and **5432** on FTK server and workers.<br>3. Check `DPE.log` in `%PROGRAMDATA%\AccessData\Logs\`. |
| **RAM Acquisition Produces Blank or Incomplete File** | Windows 11 Virtualization-Based Security (VBS) or Hypervisor-Protected Code Integrity (HVCI) blocked raw `\\Device\PhysicalMemory`. | 1. Check if HVCI is active in Windows Security $\rightarrow$ *Core Isolation*.<br>2. Use kernel-level acquisition tools (e.g. WinPmem / DumpIt driver) or suspend virtualization security during live triage.<br>3. Execute `ftkimager.exe` strictly as elevated Administrator. |
| **PostgreSQL Database Case Deadlock on Large Ingest** | Default PostgreSQL `shared_buffers` and `max_connections` parameters too low for high-core DPE clusters. | 1. In `postgresql.conf`, increase `shared_buffers = 8GB` (or 25% of RAM) and `work_mem = 64MB`.<br>2. Increase `max_connections = 300`.<br>3. Restart PostgreSQL service. |
| **OCR Processing Freezes Ingestion Pipeline** | Corrupted multi-thousand-page scanned PDF file locking OCR worker process. | 1. In Processing Profile, set OCR file size limit to $\le 50\text{MB}$.<br>2. Exclude non-document binary formats from OCR queue.<br>3. Set OCR timeout to 60 seconds per item. |

---

## Command Line Syntax & Operational Recipes

```bash
# 1. Acquire Logical Directory to AD1 Custom Content Container
ftkimager.exe "C:\Users\TargetUser\AppData" "C:\Evidence\User_AppData" --ad1 --verify

# 2. Extract Registry Hives from Live Windows System
ftkimager.exe --extract-protected-files C:\Windows\System32\config\SYSTEM C:\Evidence\SYSTEM

# 3. Mount E01 Image as Virtual Drive via FTK Imager CLI
ftkimager.exe --mount "C:\Evidence\Disk.E01" /drive=Z: /mode=read-only
```

### Essential File Locations
- **FTK Configuration**: `C:\Program Files\Exterro\FTK\FTK.cfg`
- **DPE Worker Logs**: `%PROGRAMDATA%\AccessData\Logs\DPE`
- **PostgreSQL Data Directory**: `C:\Program Files\PostgreSQL\<VER>\data`

---

## Agent Operational Directive
> **MANDATORY**: When executing FTK Imager CLI acquisitions, always include the `--verify` flag to compute and log acquisition vs verification checksums (MD5/SHA-1). Tune PostgreSQL `shared_buffers` before running high-throughput DPE ingest clusters.

Files in this skill

  • SKILL.md7.4 KB
  • gemini.md7.1 KB
  • gpt.md7.1 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.