Skip to content

vx underground Malware Research Library

github-actions[bot] edited this page Nov 15, 2026 · 1 revision

Last reviewed: 2026-10-07

Caution

Authorized use only. vx-underground hosts live malware, malware source code, and builders. Download, store, and detonate samples only inside an isolated analysis lab you control, and only for research, detection engineering, education, or authorized testing. Never run a sample on a production host, a client network, or a machine with personal data. Laws on possessing or distributing malicious code vary by jurisdiction. See LEGAL.md.

Where to get real malware, source, and research papers, and how to handle them without hurting yourself

Part of the ULTIMATE CYBERSECURITY MASTER GUIDE ยท Reverse Engineering


๐ŸŽฏ Purpose

Documents vx-underground (VXUG), a free public library of malware samples, malware source code, APT reports, and research papers, and shows how it fits into the reverse engineering, detection, and threat intel workflows already in this repo.

โš™๏ธ Function

  • Maps the site's collections and the official GitHub repositories to the job each one is good for.
  • Gives a repeatable, lab-only intake procedure: download, verify, contain, analyze, write detections.
  • Links each stage to the existing guide that covers it in depth (Ghidra, YARA, forensics, SIEM, threat intel).

๐Ÿ† Goal

Turn "I need a real sample of family X" into a safe, documented workflow that ends in something defensive: a YARA rule, a Sigma/SIEM detection, an IOC set, or a write-up.

๐Ÿ“‹ When to Use

  • You need a known sample to practice in Ghidra
  • Writing or testing YARA rules against real families
  • Validating that an EDR, AV, or SIEM rule actually fires
  • Researching an APT or ransomware group (reports + samples in one place)
  • Reading historical virus-writing and offensive research papers
  • Building a malware training set for an isolated ML or classifier project

Prerequisites: a working, isolated lab (Homelab), basic Linux CLI, and hashing/VM-snapshot habits. If you have never analyzed a binary, start with Part I of the Ghidra guide.


What vx-underground Is

  • Started in May 2019; describes itself as the largest collection of malware source code, samples, and papers on the internet.
  • Free to access, funded by sponsors and donations.
  • Runs a high-volume news/research feed on X (@vxunderground), often among the first to report ransomware leaks and breach claims.

Warning

Threat actors have impersonated vx-underground (e.g., a ransomware group in 2023 used the name to muddy attribution). Only trust content from the official domain and the official GitHub org listed below. Treat any "vx-underground" file sent to you directly as hostile.

Official channels

Channel URL Notes
Website https://vx-underground.org Main library; full-text search across file/folder names
GitHub org https://github.com/vxunderground Source code, papers, tooling
X / Twitter https://x.com/vxunderground News, leak verification, new uploads

Site Collections

Section What's in it Typical use
Samples Malware binaries, organized by family and date Ghidra practice, YARA testing, sandbox runs
APTs Vendor/government threat reports grouped by year, many with matching samples Actor profiling, ATT&CK mapping, detection gaps
Papers Offensive and defensive research papers, zines, technique write-ups Learning techniques before detecting them
Archive Historical collections (old VX groups, legacy viruses) History, classic technique study
Builders Leaked malware/ransomware builders Detection engineering only; highest-risk folder
Best Of Curated highlights Starting point when browsing
Torrents Bulk collection downloads Large offline corpora (plan storage first)

Official GitHub Repositories

Repository Purpose Use in this repo's workflows
MalwareSourceCode Malware source across many platforms and languages Read how a technique is implemented before writing detection; pair with the Checklists
VXUG-Papers Research code and papers from community members Technique deep dives; feeds Tradecraft study
VX-API C/C++ library of functions that replicate malware behavior Generate known-bad telemetry in a lab to test EDR/SIEM rules
ThreatIntelligenceDiscordBot Posts updates from clearnet sources and ransomware actor sites Self-hosted leak-site/threat feed for a SOC channel

Note

Repo contents and star counts change. Check each repo's README for current layout and license before scripting against it.


Safe Intake Workflow

1. Lab ready?      isolated VM, host-only/no network, clean snapshot taken
2. Download        from vx-underground.org or the official GitHub only
3. Move in         transfer the still-encrypted archive into the lab VM
4. Verify          hash before extraction; compare to the listing / VirusTotal
5. Extract         inside the VM only
6. Analyze         static first (strings, PE info, Ghidra), dynamic last
7. Produce         YARA / Sigma / IOCs / write-up
8. Revert          roll the VM back to the clean snapshot

1. Lab requirements

  • Dedicated analysis VM (e.g., FLARE-VM for Windows samples, REMnux for Linux tooling) with a clean snapshot. See virtualmachines.md.
  • No shared folders, no clipboard sharing, no bridged networking. Use host-only or an isolated internal network; add a fake-internet service such as INetSim on REMnux when you need dynamic network behavior.
  • Never on your daily-driver host, a client-facing machine, or a NAS/TrueNAS share that other systems mount.

2โ€“4. Download, move, verify

Keep archives encrypted until they are inside the lab. On the analysis VM:

# Hash the archive and the payload so you can track it in notes and IOC sets
sha256sum sample.7z

# List contents without extracting
7z l sample.7z

5. Extract

Packaged archives from vx-underground "may or may not" use the industry-standard password infected, per the MalwareSourceCode README. Individual unpacked files usually have no password.

# Extract into a dedicated working folder inside the lab VM
mkdir -p ~/cases/<case-id>/raw
7z x -p'infected' sample.7z -o"$HOME/cases/<case-id>/raw"

# Hash every extracted file
find ~/cases/<case-id>/raw -type f -exec sha256sum {} + | tee ~/cases/<case-id>/hashes.txt

Tip

When you move a sample out of the lab (to a colleague or another VM), re-zip it with the password infected and rename the extension (e.g. .exe_) so nothing auto-executes it and mail/AV gateways do not silently strip it.

6. Analyze

Stage Where in this repo
File triage, strings, imports Ghidra guide, Part Iโ€“II
Packed / obfuscated samples Ghidra guide, Part V
Headless batch analysis of many samples Ghidra guide, Part III
Memory artifacts after detonation Digital Forensics
Host telemetry while detonating Endpoint Visibility

7. Produce something defensive

  • YARA: write the rule against the sample, then test it against a clean corpus to check false positives.
  • SIEM / Sigma: detonate with Sysmon running and build detections from the telemetry; see SIEM.
  • IOCs: hashes, C2 domains, mutexes, registry keys; push them into MISP or OpenCTI (see OSINT & Threat Intel).
  • ATT&CK mapping: tag behaviors with MITRE ATT&CK technique IDs so the write-up lines up with the APT reports.

Using the APT Collection for Threat Intel

  1. Pick the actor or campaign relevant to your sector.
  2. Read the vendor reports in the APT folder for that year.
  3. Pull listed hashes, then grab the matching samples (from VXUG or MalwareBazaar).
  4. Map observed behaviors to ATT&CK and compare against your current detections.
  5. Close gaps, then retest with the sample or with VX-APIโ€“generated telemetry.

This is the same intelligence cycle described in osint-threat-intel.md, with real samples as the collection source.


Complementary Sample Sources

Source Strength
MalwareBazaar (abuse.ch) Fresh, tagged samples with API and hash exports
VirusShare Very large hash sets and bulk archives (account required)
VirusTotal Reputation, relationships, retrohunt (sample download requires paid tiers)
theZoo Small, curated live-malware repo for teaching

Risks and Rules

Caution

  • Builders and VX-API produce working malicious capability. Use them only to generate lab telemetry. Never compile or deploy their output outside the lab, and never use them in a client engagement without a written scope that explicitly covers it.
  • Some samples are wormable or destructive (ransomware, wipers). Assume every sample will try to spread and encrypt anything it can reach.
  • Samples in your possession may be subject to client NDAs, export rules, or local computer-misuse law. Document why you hold each sample.
  • Disable cloud sample submission in any AV you run in the lab unless you intend to share the file.

References

See also


โฌ…๏ธ Back to Master Index | ๐ŸŽฏ Role Navigation | Legal Notice


๐Ÿ“„ Source: ReverseEngineering/vx-underground.md | โœ๏ธ Suggest an edit | โฌ…๏ธ Back to Reverse Engineering

Clone this wiki locally