OSINT Techniques: Chapter 1 - Prepare Your Computer
An exhaustive breakdown of Chapter 1 of Michael Bazzell's 'Open Source Intelligence Techniques', analyzing workstation hygiene, browser hardening, anti-fingerprinting, and robust sock puppet construction.
Welcome to the foundational post of our exhaustive book summary series covering “Open Source Intelligence Techniques: Resources for Searching and Analyzing Online Information” (Sixth Edition) by former FBI Computer Crimes Task Force investigator Michael Bazzell.
In offensive security, cyber threat intelligence, and law enforcement investigations, open source intelligence represents the initial, decisive phase of engagement. Before an analyst queries a target’s infrastructure, explores social media footprints, or interacts with web services, workstation operational security (OpSec) dictates whether the investigation remains strictly covert or inadvertently alerts the target.
In Chapter 1: Prepare Your Computer, Bazzell establishes the critical axiom of digital investigation: your personal computer, default browser, and everyday internet identity are toxic to investigative integrity. Every interaction leaves digital breadcrumbs—IP addresses, browser canvas fingerprints, TLS handshake signatures, WebRTC leaks, and behavioral telemetry. In this deep dive, we deconstruct workstation preparation, deep browser hardening, network isolation, and the end-to-end lifecycle of investigative personas (sock puppets).
Key Takeaways
- The Attribution Paradox: Browsing the web in default configurations broadcasts hundreds of unique hardware and software identifiers. Operating without isolation guarantees cross-contamination between personal identity and investigative targets.
- Layered Defense-in-Depth: True investigative security requires separation at the physical hardware, operating system, network egress, and browser application levels.
- Browser Fingerprinting Dominance: Modern anti-fraud and tracking systems identify clients not through cookies or IP addresses alone, but through canvas rendering noise, AudioContext signatures, WebGL vendor strings, and installed font enumeration.
- WebRTC as an OpSec Killer: WebRTC enables peer-to-peer browser communication but bypasses standard application-level HTTP proxies, leaking real local and public IP addresses via STUN/TURN requests unless explicitly disabled at the engine level.
- Sock Puppet Engineering: A viable sock puppet is an engineered digital identity, requiring pre-aging, authentic non-VoIP SMS verification, separate masked financial instruments, and strictly compartmentalized browsing profiles.
Architectural & Investigative Workflow
The following flowchart outlines the multi-tiered isolation model required for non-attributable digital investigations:
flowchart TD
subgraph Physical_Hardware["Layer 1: Physical Hardware & OS"]
Hardware["Dedicated Non-Attributable Laptop / Clean Drive"] --> OS["Hardened Host OS (Linux Debian/Qubes / macOS)"]
OS --> FDE["Full Disk Encryption (LUKS2 / FileVault)"]
FDE --> Telemetry["Disabled System Telemetry & Macchanger"]
end
subgraph Network_Boundary["Layer 2: Network Egress & Routing"]
Telemetry --> FW["Host Firewall (UFW Egress Filtering)"]
FW --> Tunnel["WireGuard Multi-Hop VPN / Tor SOCKS5 Gateway"]
Tunnel --> DNS_Sec["Encrypted DNS-over-HTTPS (DoH) / Zero Leaks"]
end
subgraph Browser_Sandboxing["Layer 3: Browser Sandboxing"]
DNS_Sec --> Profile["Firefox Custom user.js / LibreWolf"]
Profile --> Containers["Firefox Multi-Account Containers"]
Containers --> Canvas["CanvasBlocker & Anti-Fingerprinting"]
Containers --> UBO["uBlock Origin (Hard Mode Blocking)"]
end
subgraph Operational_Identity["Layer 4: Persona (Sock Puppet)"]
UBO --> Sock["Aged Sock Puppet Identity"]
Sock --> Verification["Real SIM/eSIM + Masked Virtual Card"]
end
Technical Breakdown & Investigative Mechanics
1. Workstation Hardware & Host Operating System Hygiene
Operating from a daily personal laptop introduces insurmountable cross-contamination risks. A single background service (such as Google Drive, OneDrive, Slack, or Apple iCloud) syncing during an investigation can reveal the investigator’s true IP or hardware UUIDs.
Hardware Isolation Rules
- Dedicated Machine: Utilize a dedicated laptop purchased with cash or sanitized secondary hardware devoid of personal accounts or biometric enrollment (Touch ID / Windows Hello).
- Full Disk Encryption (FDE): Enforce AES-XTS 256-bit LUKS2 encryption at rest. If the device is seized or physically inspected, investigative caches, browser histories, and downloaded target evidence remain cryptographically inaccessible.
- Hardware MAC Randomization: Physical Network Interface Cards (NICs) carry immutable vendor Organizationally Unique Identifiers (OUIs). Before connecting to any investigative network or Wi-Fi hotspot, randomize the MAC address:
1
2
3
4
5
6
7
8
9
10
# Verify current MAC address
ip link show wlan0
# Bring interface down, randomize MAC, and restore interface
sudo ip link set wlan0 down
sudo macchanger -r wlan0
sudo ip link set wlan0 up
# Verify randomized MAC assignment
macchanger -s wlan0
2. Deep Browser Hardening & Custom user.js Configuration
Standard commercial browsers (Google Chrome, Microsoft Edge) are engineered to maximize user tracking, behavioral telemetry, and persistent session identification. For OSINT research, Mozilla Firefox (or dedicated forks like LibreWolf) configured with custom user.js directives represents the gold standard.
Critical Firefox Configuration Directives
Placing a custom user.js inside the Firefox profile folder (~/.mozilla/firefox/<profile_name>/user.js) locks privacy settings against automatic browser updates or accidental UI toggles:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
// user.js - High-Security OSINT Hardening Configuration
// 1. Disable WebRTC to prevent STUN/TURN IP address leakage
user_pref("media.peerconnection.enabled", false);
user_pref("media.peerconnection.use_document_iceservers", false);
// 2. Enable Mozilla Resist Fingerprinting (RFP) engine
// Spoofs User-Agent, locks screen resolution to 1000x1000 / 1200x800, sets UTC timezone
user_pref("privacy.resistFingerprinting", true);
user_pref("privacy.resistFingerprinting.letterboxing", true);
// 3. Disable WebGL and Canvas Reading APIs
user_pref("webgl.disabled", true);
user_pref("webgl.enable-webgl2", false);
// 4. Enforce strict cookie isolation and first-party isolation
user_pref("network.cookie.cookieBehavior", 1); // Reject 3rd-party cookies
user_pref("privacy.firstparty.isolate", true);
// 5. Disable Browser Telemetry, Health Reports, and Crash Reporter
user_pref("datareporting.healthreport.uploadEnabled", false);
user_pref("toolkit.telemetry.enabled", false);
user_pref("toolkit.telemetry.unified", false);
user_pref("experiments.supported", false);
// 6. Force Encrypted DNS-over-HTTPS (DoH) via Cloudflare or Quad9
user_pref("network.trr.mode", 3); // Mode 3 = Exclusive DoH, fail closed
user_pref("network.trr.uri", "https://dns.quad9.net/dns-query");
3. Extension Architecture & Anti-Fingerprinting Defense
While extensions enhance security, installing dozens of obscure add-ons paradoxically creates a unique browser fingerprint via chrome.runtime resource enumeration. An investigator must maintain a lean, high-impact extension stack:
| Extension Name | Operational Purpose | Mandatory Configuration |
|---|---|---|
| uBlock Origin | Blocking telemetry, third-party frames, trackers | Switch to Medium / Hard Mode; block 3rd-party scripts and inline frames by default |
| CanvasBlocker | Spoofing HTML5 Canvas, AudioContext, DOMRect | Set to Fake Readout; injects subtle persistent random noise per domain |
| Multi-Account Containers | Contextual identity isolation | Separate containers for: Personal, Target A, Target B, and Sock Puppet |
| User-Agent Switcher and Manager | Emulating common platforms | Match the User-Agent to the host architecture to prevent font-to-UA discrepancies |
4. Network Layer: Multi-Hop VPNs, Tor & DNS Leak Remediation
Network egress must guarantee that the target web server never observes the investigator’s residential or enterprise IP address. Furthermore, DNS resolution must never travel unencrypted to a local Internet Service Provider (ISP).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# Configure Linux UFW (Uncomplicated Firewall) Kill-Switch for WireGuard (wg0)
sudo ufw default deny incoming
sudo ufw default deny outgoing
# Allow loopback interface
sudo ufw allow in on lo
sudo ufw allow out on lo
# Allow outbound traffic ONLY through WireGuard VPN interface
sudo ufw allow out on wg0 to any
# Allow initial VPN handshake to the WireGuard server endpoint
sudo ufw allow out to 198.51.100.25 port 51820 proto udp
# Enable firewall and verify status
sudo ufw enable
sudo ufw status verbose
Verifying Zero-Leak Status via Terminal
Before initiating any investigation, run an automated network verification pipeline:
1
2
3
4
5
6
#!/usr/bin/env bash
echo "[*] Auditing Public IP & Geolocation..."
curl -s --max-time 5 https://ipinfo.io/json | jq '{ip: .ip, city: .city, org: .org}'
echo "[*] Testing for WebRTC / DNS Leakage..."
dig +short whoami.akamai.net @208.67.222.222
5. Sock Puppet Engineering & Operational Persona Lifecycle
A Sock Puppet is a fictitious online identity designed to conduct investigations on authenticated platforms (Facebook, LinkedIn, Instagram, TikTok, Telegram) without alerting the target or revealing personal affiliations.
sequenceDiagram
autonumber
participant Inv as Investigator
participant Mint as Identity Minting
participant Tel as Telephony / SMS
participant Plat as Social Media Platform
participant Mon as Long-Term Maintenance
Inv->>Mint: Generate Persona Details (Name, Age, Location, Background)
Inv->>Tel: Provision Real Non-VoIP SIM (Physical SIM / Silent.link)
Inv->>Plat: Register Account via Dedicated Container & Fixed Residential VPN
Plat-->>Tel: Dispatch SMS 2FA Code
Tel-->>Inv: Relay 2FA Token
Inv->>Plat: Confirm Registration & Complete Non-Target Onboarding
Inv->>Mon: 30-Day "Aging" Phase: Join Local Groups, Add Distant Contacts
Mon-->>Plat: Natural Browsing Activity (Prevents Algorithmic Shadowban)
Sock Puppet Construction Rules
- The Human Realism Test: Do not use AI-generated images from ThisPersonDoesNotExist.com as profile pictures on modern platforms. Deep learning detection models analyze pupil symmetry, corneal reflections, and hair artifacting to instantly flag new accounts. Instead, use real stock photos altered with subtle filters, or high-angle scenic photos devoid of facial recognition vectors.
- True Non-VoIP Telephone Numbers: Automated fraud engines (Telesign, Twilio Lookup) immediately detect Google Voice, TextNow, or Skype numbers. Acquire prepaid burner SIM cards registered with cash, or specialized privacy eSIM providers (e.g., Silent.link, Mint Mobile) that return legitimate carrier Type-Allocation Codes (TAC).
- Dedicated Residential IP Pools: Never access social media sock puppets via public Tor exit nodes or datacenter VPN IPs (DigitalOcean, AWS, OVH). Platforms track Autonomous System Numbers (ASNs) and flag datacenter traffic immediately. Utilize fixed residential VPN proxies or 4G/5G mobile proxies.
- The 30-Day Aging Protocol: Newly registered accounts that immediately search for specific targets, join private groups, or view hundreds of profiles are instantly terminated. Age the sock puppet for 3-4 weeks: like local news pages, follow regional sports teams, and browse benign content to build a trustworthy algorithmic baseline.
Real-World Case Scenario
Case Study: Counter-Intelligence Infiltration Assessment
An investigator was tasked with analyzing an organized threat group operating a network of fraudulent cryptocurrency investment schemes across Telegram and private Facebook groups.
An amateur investigator used a fresh Google Chrome profile without disabling WebRTC. The threat group operated a custom landing page embedded in their group link. When the analyst visited the page, WebRTC executed a silent STUN lookup, bypassing the commercial VPN and logging the analyst’s real corporate IP address. The threat group immediately blocked the account and deleted their Telegram channels.
In contrast, an investigator following Chapter 1 protocols accessed the infrastructure via a hardened Firefox instance with
media.peerconnection.enabled = false, egressing through an isolated WireGuard kill-switch tunnel and an aged sock puppet. The group’s tracking scripts received spoofed canvas noise, a generic Windows/Firefox User-Agent, and zero real IP leakage, allowing the investigator to infiltrate the group, extract member rosters, and archive incriminating evidence undetected.
Defensive Countermeasures & OpSec Recommendations
For corporate security teams, executives, and high-risk individuals seeking to minimize their online exposure:
- Regular Browser Fingerprint Audits: Test corporate workstations using tools like
Cover Your Tracks(EFF) andBrowserleaks.comto identify unmitigated tracking vectors. - Disable WebRTC Across Enterprise Fleets: Deploy central Group Policy Objects (GPOs) or mobile device management (MDM) profiles disabling WebRTC on browsers not requiring browser-based VOIP/video conferencing.
- Monitor Inbound Sock Puppet Reconnaissance: Inspect corporate LinkedIn and social media visitor logs for accounts exhibiting classic sock puppet traits: single company followers, zero mutual educational institutions, recently minted profiles, and stock-style profile images.
Conclusion & Next Steps
Chapter 1 underscores that investigative success is fundamentally tethered to workstation discipline. A compromised workstation renders even the most advanced search techniques dangerous to the investigator and useless in court.
In Chapter 2: Linux Investigation VMs, we move from host hardening to architecting dedicated, disposable Linux virtual environments, exploring the legacy of Buscador, modern distributions like CSI Linux and Trace Labs OSINT VM, and automated hypervisor isolation workflows.
Practical Workstation Hardening & OpSec Cheat-Sheet
| Operational Control | Configuration / Command | Security Objective |
|---|---|---|
| Disable WebRTC IP Leaks | media.peerconnection.enabled = false in Firefox about:config |
Prevents STUN lookups from leaking real local and public IP addresses past VPNs |
| Canvas & Font Spoofing | privacy.resistFingerprinting = true in Firefox about:config |
Injects synthetic noise into HTML5 canvas renders and normalizes system font enumeration |
| Verify Egress IP Address | curl -s https://api.ipify.org & curl -s https://ipinfo.io/json |
Confirms outbound traffic routes strictly through the designated VPN/proxy gateway |
| Audit STUN Leakage | tshark -i any -f "udp port 3478 or udp port 19302" |
Sniffs local network interfaces for unauthorized STUN/TURN traversal packets |
| DNS Leak Verification | dig +short whoami.akamai.net |
Validates that recursive DNS queries originate from the VPN resolver, not local ISP |
Authoritative References & Verified Investigative Sources
- Michael Bazzell, IntelTechniques: Open Source Intelligence Techniques: Resources for Searching and Analyzing Online Information (6th Edition). https://inteltechniques.com/book1.html.
- Electronic Frontier Foundation (EFF): Cover Your Tracks - Browser Fingerprinting & Tracking Protection Auditing. https://coveryourtracks.eff.org/.
- BrowserLeaks Research: WebRTC Leaks, Canvas Fingerprinting & Hardware Concurrency Testing. https://browserleaks.com/webrtc.
- Bellingcat Digital Forensics: Guides to Online Research, Digital Security, and Verification. https://www.bellingcat.com/resources/.
- NIST Special Publication 800-61 Rev. 2: Computer Security Incident Handling Guide. https://csrc.nist.gov/publications/detail/sp/800-61/rev-2/final.
Author: Abdulrahman M. Fathi (0x3bdo) • Cyber Security Engineer & Penetration Tester