Skip to content

Various configuration files for Microsoft Windows operating systems

License

Notifications You must be signed in to change notification settings

straysheep-dev/windows-configs

Repository files navigation

windows-configs

Various configuration settings and notes for Microsoft Windows operating systems.

Licenses

Unless a different license is included with a file as <filename>.copyright-notice all files are released under the MIT license.

Examples in this README taken and adapted from the Microsoft documents:

Windows Baselining

Creating a security baseline for Windows.

Backup the Registry

Before making changes, it's useful to create a backup of the registry in a default or working state.

To create a system restore point:

Enable-ComputerRestore -Drive "C:\"
Checkpoint-Computer -Description "First restore point"

There are multiple restore point types:

  • APPLICATION_INSTALL
  • APPLICATION_UNINSTALL
  • DEVICE_DRIVER_INSTALL
  • MODIFY_SETTINGS
  • CANCELLED_OPERATION

For registry changes specifically, APPLICATION_INSTALL is the type to use. It's also the default type, so it's not necessary to specify it on the commandline.

List all system restore points:

Get-ComputerRestorePoint

After making changes, revert to a specific restore point:

Restore-Computer -RestorePoint 1

The restore process can take several minutes, even when reverting a single change to the registry. Generally it takes about 3-4 minutes for local test VM's.

Applying a Baseline

See the tools available in the Microsoft Security Compliance Toolkit

Choose what tools and policies to download that you'd like to apply to your environment.

The idea with the .PolicyRules files is they are configurations that are pre-made by Microsoft and ready to be installed using LGPO.exe

You can do all of this manually with PowerShell, and you will ultimately want to familiarize yourself with the descriptions of each setting should you run into any issues, but this will save a ton of time in getting things up and running.

Use PolicyAnalyzer.exe to view the *.PolicyRules files, compare them to other *.PolicyRules files or even your current system settings.

Use LGPO.exe to apply the configurations found in the *.PolicyRules files to your system.

For example the you might apply the Security Baselines for both Windows 11 Pro and the latest available version of Microsoft Edge.

Deploy and test these configurations in a temporary or virtual environment first, either a VM (local or cloud) or enabled the Windows Sandbox feature.

Windows Sandbox is a temporary, and (depending on your .wsb configuration) fully isolated environment that can be started very quickly from either launching the application as you would any other, or by running a .wsb configuration file.

Active Directory

AD: Installing the AD RSAT Tools

You no longer need to download the AD RSAT packages from Microsoft's website, instead these are included as Windows Features on Demand. The AD RSAT tools come available by default in most Windows Server installations. However you can add the capability manually to a workstation by following the references below.

You do not need to enable and install every feature to start working with AD. Some essentials, considering this is likely being added to a workstation or server (separate from the DC):

  • Active Directory Domain Services and Lightweight Directory Services Tools
  • Active Directory Certificate Services Tools
  • Group Policy Management Tools

To install via the GUI, search for "RSAT" then select the features you want to install:

  • Settings > Apps > Optional Features > View Features > Search "RSAT"

To install via the CLI, first get a list of all available features then install the features you want by name:

DISM /online /get-capabilities
DISM /online /add-capability /capabilityname:<capability-name>
DISM /online /add-capability /capabilityname:Rsat.ActiveDirectory.DS-LDS.Tools~~~~0.0.1.0

AD: RSAT Tool Usage

The cmdlet syntax is similar to other PowerShell cmdlets. PowerView has great examples of how you can enumerate AD objects, and uses mostly the same syntax as the AD RSAT modules. PowerView is the tool to enumerate AD if you do not have the AD RSAT modules installed on an end point and lack permissions to install them. Note that it will need allowed by AV / EDR.

AD: Group Policy

Update Group Policy to sync with the Domain Controller via cmd.exe:

gpupdate.exe

Push updates to Group Policy to sync with the Domain Controller via PowerShell:

Invoke-GPUpdate -Computer "DOMAIN\COMPUTER-01" -Target "User" -RandomDelayInMinutes 0 -Force

NOTE: some GPO's do not appear configured in the gpedit.msc snap in on endpoints, but will appear if you check their registry.

One example of this is Disabling Link-Local Multicast Name Resolution (LLMNR) protocol.

Setting Computer Configuration > Administrative Templates > Network > DNS Client > Turn off multicast name resolution to Enabled on the Domain Controller and pushing the update with Invoke-GPUpdate or similar will not display this policy as Enabled on the endpoints, however it will show as set in their registry. This can be tested by changing the policy on the DC and refreshing each endpoint before checking again with Get-ItemProperty -Path "HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\DNSClient".

Managing Windows Security

Configuring Device Lock

Require authentication after resuming from sleep (this should be on by default):

  • Local Computer Policy > User Configuration > Administrative Templates > System > Power Management > "Prompt for password on resume from hibernate/suspend"
  • You only need to run this as admin to apply this setting system-wide
If (!(Test-Path "HKCU:\SOFTWARE\Policies\Microsoft\Windows\System\Power")) {
	New-Item -Path "HKCU:\SOFTWARE\Policies\Microsoft\Windows\System\Power" -Force | Out-Null
}
Set-ItemProperty -Path "HKCU:\SOFTWARE\Policies\Microsoft\Windows\System\Power" -Name "PromptPasswordOnResume" -Type DWord -Value "1"

Require authentication (immediately) after screen turns off:

  • This has no GPO equivalent
  • This also appears to have no GUI configuration option either
  • This setting must be configured separately for each user, including the admin account
Set-ItemProperty -Path "HKCU:\Control Panel\Desktop" -Name "DelayLockInterval" -Type DWord -Value "0"

Windows Defender Cmdlets

https://docs.microsoft.com/en-us/powershell/module/defender/?view=windowsserver2022-ps

Cmdlet Description
Add-MpPreference Modifies settings for Windows Defender.
Get-MpComputerStatus Gets the status of antimalware software on the computer.
Get-MpPreference Gets preferences for the Windows Defender scans and updates.
Get-MpThreat Gets the history of threats detected on the computer.
Get-MpThreatCatalog Gets known threats from the definitions catalog.
Get-MpThreatDetection Gets active and past malware threats that Windows Defender detected.
Remove-MpPreference Removes exclusions or default actions.
Remove-MpThreat Removes active threats from a computer.
Set-MpPreference Configures preferences for Windows Defender scans and updates.
Start-MpScan Starts a scan on a computer.
Start-MpWDOScan Starts a Windows Defender offline scan.
Update-MpSignature Updates the antimalware definitions on a computer.

Controlled Folder Access

Protect your data from malicious apps such as ransomware

Write access must be granted to applications before modifications can be made to files within folders you define as protected. This is often the home directories under C:\Users, but can also be filesystems on external drives.

Enable Controlled Folders

https://docs.microsoft.com/en-us/microsoft-365/security/defender-endpoint/enable-controlled-folders?view=o365-worldwide

Set-MpPreference -EnableControlledFolderAccess Enabled
Set-MpPreference -EnableControlledFolderAccess AuditMode # Enable the audit feature only
Set-MpPreference -EnableControlledFolderAccess Disabled  # Turn off Controlled Folder Access

Customize Controlled Folder Access

https://docs.microsoft.com/en-us/microsoft-365/security/defender-endpoint/customize-controlled-folders?view=o365-worldwide

IMPORTANT: Use Add-MpPreference to append or add apps to the list and not Set-MpPreference. Using the Set-MpPreference cmdlet will overwrite the existing list.

Add/remove protection for a folder:

Add-MpPreference -ControlledFolderAccessProtectedFolders "c:\path\to\folder"
Remove-MpPreference -ControlledFolderAccessProtectedFolders "c:\path\to\folder"

Add/remove an application's access to protected folders:

Add-MpPreference -ControlledFolderAccessAllowedApplications "c:\apps\test.exe"
Remove-MpPreference -ControlledFolderAccessAllowedApplications "c:\apps\test.exe"

ASR (Attack Surface Reduction)

Rule Name Rule GUID
Block abuse of exploited vulnerable signed drivers 56a863a9-875e-4185-98a7-b882c64b5ce5
Block Adobe Reader from creating child processes 7674ba52-37eb-4a4f-a9a1-f0f9a1619a2c
Block all Office applications from creating child processes d4f940ab-401b-4efc-aadc-ad5f3c50688a
Block credential stealing from the Windows local security authority subsystem (lsass.exe) 9e6c4e1f-7d60-472f-ba1a-a39ef669e4b2
Block executable content from email client and webmail be9ba2d9-53ea-4cdc-84e5-9b1eeee46550
Block executable files from running unless they meet a prevalence, age, or trusted list criterion 01443614-cd74-433a-b99e-2ecdc07bfc25
Block execution of potentially obfuscated scripts 5beb7efe-fd9a-4556-801d-275e5ffc04cc
Block JavaScript or VBScript from launching downloaded executable content d3e037e1-3eb8-44c8-a917-57927947596d
Block Office applications from creating executable content 3b576869-a4ec-4529-8536-b80a7769e899
Block Office applications from injecting code into other processes 75668c1f-73b5-4cf0-bb93-3ecf5cb7cc84
Block Office communication application from creating child processes 26190899-1602-49e8-8b27-eb1d0a1ce869
Block persistence through WMI event subscription (File and folder exclusions not supported). e6db77e5-3df2-4cf1-b95a-636979351e5b
Block process creations originating from PSExec and WMI commands d1e49aac-8f56-4280-b9ba-993a6d77406c
Block untrusted and unsigned processes that run from USB b2b3f03d-6a65-4f7b-a9c7-1c7ef74a9ba4
Block Win32 API calls from Office macros 92e97fa1-2edf-4476-bdd6-9dd0b4dddc7b
Use advanced protection against ransomware c1db55ab-c21a-4637-bb3f-a12568109d35

ASR rules can be set in multiple modes:

  • 0 = Not configured / Disabled
  • 1 = Block
  • 2 = Audit
  • 6 = Warn

Set a rule by it's GUID with PowerShell:

Add-MpPreference -AttackSurfaceReductionRules_Ids <guid> -AttackSurfaceReductionRules_Actions <mode>

To enable all rules at once:

Set-MpPreference -AttackSurfaceReductionRules_Ids 56a863a9-875e-4185-98a7-b882c64b5ce5 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 7674ba52-37eb-4a4f-a9a1-f0f9a1619a2c -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids d4f940ab-401b-4efc-aadc-ad5f3c50688a -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 9e6c4e1f-7d60-472f-ba1a-a39ef669e4b2 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids be9ba2d9-53ea-4cdc-84e5-9b1eeee46550 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 01443614-cd74-433a-b99e-2ecdc07bfc25 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 5beb7efe-fd9a-4556-801d-275e5ffc04cc -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids d3e037e1-3eb8-44c8-a917-57927947596d -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 3b576869-a4ec-4529-8536-b80a7769e899 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 75668c1f-73b5-4cf0-bb93-3ecf5cb7cc84 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 26190899-1602-49e8-8b27-eb1d0a1ce869 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids e6db77e5-3df2-4cf1-b95a-636979351e5b -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids d1e49aac-8f56-4280-b9ba-993a6d77406c -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids b2b3f03d-6a65-4f7b-a9c7-1c7ef74a9ba4 -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids 92e97fa1-2edf-4476-bdd6-9dd0b4dddc7b -AttackSurfaceReductionRules_Actions 1
Add-MpPreference -AttackSurfaceReductionRules_Ids c1db55ab-c21a-4637-bb3f-a12568109d35 -AttackSurfaceReductionRules_Actions 1

Confilcting Policy

If a conflicting policy is applied via MDM and GP, the setting applied from MDM will take precedence.

Exclude files and folders from ASR rules

You can specify individual files or folders (using folder paths or fully qualified resource names), but you can't specify which rules the exclusions apply to.

Excluding files and folders from ASR rules with PowerShell

Add-MpPreference -AttackSurfaceReductionOnlyExclusions "<fully qualified path or resource>"

Windows Sandbox

Example 1 provides a great base configuration for a malware analysis setup, where GPU and Networking are disabled, and a single folder from the host is available as read-only.

Example 2 demonstrates mapping different host folders to the sandbox as read-only or writable, and also reading from a cmd script on the host to execute on startup, which downloads and installs VSCode automatically.

The document also notes exposing the following features to the sandbox potentially affects the attack surface:

  • vGPU (Disabled by default)
  • Network (Enabled by default)
  • Mapped folders and files that are writable
  • Audio input (Enabled by default)
  • Video input (Disabled by default)

There's also an option called Protected Client mode:

"Applies more security settings to the sandbox Remote Desktop client, decreasing its attack surface."

This repo contains a .wsb configuration file geared towards malware analysis that follows Example 1, and disables / enables a few additional features. Without networking in the sandbox, you should plan to have installers for all of the required tools within a mapped folder (C:\Tools -> C:\Users\WDAGUtilityAccount\Documents\Tools). What you can do then is save a .cmd script within the mapped Tools directory to launch with as many install commands as you're able to automate:

So if the script is named install.cmd, the .wsb file contains these lines:

  <LogonCommand>
    <Command>C:\Users\WDAGUtilityAccount\Documents\Tools\install.cmd</Command>
  </LogonCommand>

And install.cmd itself could look something like this:

C:\Users\WDAGUtilityAccount\Documents\Tools\tool1.exe /install
C:\Users\WDAGUtilityAccount\Documents\Tools\tool2.exe /arg 1 /arg 2 /install
C:\Users\WDAGUtilityAccount\Documents\Tools\tool3.exe /install

If you wish to automate any tasks using PowerShell, you can do so by having a .cmd script set the execution policy for PowerShell, then execute any number of .ps1 scripts (which could also contain references to other PowerShell scripts). So the .cmd script could look like the following:

cmd.exe /C powershell.exe -c Set-ExecutionPolicy Bypass -Force
cmd.exe /C powershell.exe C:\Users\WDAGUtilityAccount\Documents\Tools\SandboxSetup.ps1

SandboxSetup.ps1 as an example could contain the following to import modules and execute those functions with arguments:

Import-Module C:\Users\WDAGUtilityAccount\Documents\Tools\Set-EdgePolicy.ps1
Set-EdgePolicy Apply

Windows Sandbox + VPN

Windows Sandbox is limited in networking options when compared to Hyper-V VMs. However, it's lightweight and highly configurable, making it easy to network using something like Wireguard.

Two interesting use cases as examples, the first is detailed in the link:

The second uses the same idea, but with a public VPN provider that can generate disposable Wireguard configurations. The keyword being disposable in that the configuration does not reveal your username, password, or account data tied to the VPN provider, and the configuration can be revoked manually at any time by you. Assuming during an OSINT investigation Windows Sandbox could be compomised, you would not want those details stolen.

To install the latest Wireguard Windows client:

$progressPreference = 'silentlyContinue'
cd $env:TEMP; iwr https://download.wireguard.com/windows-client/wireguard-installer.exe -OutFile wireguard-installer.exe; .\wireguard-installer.exe

Wireguard will open once installed.

  • You can go to Add Tunnel > Add empty tunnel... or Ctrl+n to open a blank configuration
  • Generate a client configuration using your VPN provider
  • Copy and paste that block of text into the large text field, replacing the default [Interface] and PrivateKey data.
  • Check Block untunneled traffic (kill-switch) if needed
  • Save

Activate the tunnel, then run the following to verify your public IP:

curl.exe https://ipinfo.io/ip

When you're done, deactivate the tunnel and revoke the client configuration using your VPN provider.

Some things to keep in mind:

  • Windows Sandbox has no firewall rules, any listening services on all interfaces will be reachable by other VPN clients if the VPN is untrusted
  • DNS should be forwarded over Wireguard unless you're configuring an alternative way to do DNS in the .wsb file / sandbox

Tailscale works similarly, and also has a "latest" executable installer for convenience:

$progressPreference = 'silentlyContinue'
cd $env:TEMP; iwr https://pkgs.tailscale.com/stable/tailscale-setup-latest.exe -OutFile tailscale-setup-latest.exe; .\tailscale-setup-latest.exe

Walk through the installer, when finished authenticate to a tailnet using PowerShell with:

tailscale.exe up --authkey tskey-<your-key-here>

Windows Sandbox Limitations

  • Docker on Windows requires Hyper-V containerization or WSL
  • Python libraries may fail to load missing DLLs
    • This is true if installing Google's magika even via pip or pipx, onnxruntime may fail to find a missing DLL

WSL

Windows Subsystems for Linux.

Install

The above article covers everything to get WSL running on your machine. The new wsl --install command installs WSL 2 by default.

If you previously installed WSL (v1) and need to upgrade, see the following resources:

Systemd support was added to WSL and is enabled by default. This will allow for things like snap, systemctl, dns daemons, and other things to be installed and used on WSL.

Even if you installed WSL2 recently, you should be sure to check your version information.

Update WSL(2) to the latest release if you're still missing systemd functionality:

wsl --update      # Update WSL
wsl --shutdown    # Restart WSL
wsl               # Open a new shell

If your wsl instance still does not have systemd running, check /etc/wsl.conf, create it if it doesn't exist, and make sure it has the following lines:

[boot]
systemd=true

Once it does, exit wsl and run wsl --shutdown; wsl to restart wsl with systemd.

Configure

Adjust WSL settings, using wsl.conf to set a hostname, the default user, or with .wslconfig to change automount policies, whether to use Windows Firewall for WSL, or limit host RAM usage.

Below are some settings that are often used, when you need a unique hostname, systemd enabled, and you're using your own DNS daemon such as unbound or stubby instead of systemd-resolved:

# wsl.conf

[boot]
systemd=true

[network]
hostname = WSL-USER
generateResolvConf = false

[user]
default=myuser

Communicating with Hyper-V

WSL's vEthernet (WSL (Hyper-V firewall)) and Hyper-V's vEthernet (Default Switch) are two separate subnets running virtually on your host. By default Windows blocks all inbound traffic that doesn't have an explicit allow rule. You can write allow rules, however this doesn't work well with these virtual interfaces as they're regenerated with new information on reboot. Instead the more robust option is to configure these two adapters to communicate with each other. Regardless of the network information, the adapter names tend to stay the same unless there is an update in Windows that changes them for some reason (this happened with WSL's adapter name, changing it from vEthernet (WSL) to vEthernet (WSL (Hyper-V firewall))). This all happens internally on your host.

# Apply
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (WSL (Hyper-V firewall))' -or $_.InterfaceAlias -eq 'vEthernet (Default Switch)'} | Set-NetIPInterface -Forwarding Enabled

# Remove
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (WSL (Hyper-V firewall))' -or $_.InterfaceAlias -eq 'vEthernet (Default Switch)'} | Set-NetIPInterface -Forwarding Disabled

NOTE: Windows does not allow (drops) ICMP reply packets by default. Try connecting to the Hyper-V VM's service directly from WSL. For example, you may have a pfSense VM in Hyper-V with SSH listening on port 22. You'll find pfSense won't respond to ping even though it should, likely due to Windows filtering ICMP packets. If you nmap -n -Pn -sT -p22 -e eth0 --open HYPER_V_PFSENSE_IP you'll find the port is open.

Keep in mind if you want to make a service running on WSL2 available to other (external) networks, you'll need to port forward the connection.

External Inbound Connections

Using a python3 web server running on WSL as an example, you can copy and paste the following code snippet to configure the Windows host to allow and forward incoming connections to the python web server.

To make this work:

  • A firewall rule allowing the connection on the specified $http_port
  • A netsh portproxy rule forwarding the connection from Windows to the WSL IP where the webserver is listening
$http_port = "8080"
$win_ipv4 = Get-NetIPAddress -InterfaceAlias "Ethernet*" -AddressFamily IPv4 | Select -First 1 IPAddress | ForEach-Object { $_.IPAddress }
$wsl_ipv4 = wsl.exe ip addr show eth0 | sls "(?:(?:192\.168|172\.16|10\.(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?))\.)(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?\.)(?:25[0-4]|2[0-4][0-9]|[01]?[0-9][0-9]?)" | ForEach-Object { $_.Matches.Value }

# Add the rules and connection
netsh interface portproxy add v4tov4 listenport="$http_port" listenaddress=$win_ipv4 connectport="$http_port" connectaddress=$wsl_ipv4
New-NetFirewallRule -DisplayName "WSL Portproxy" -Profile Any -Direction Inbound -Protocol TCP -LocalPort "$http_port"
wsl.exe python3 -m http.server "$http_port" --bind "$wsl_ipv4"

Shut down the server, then remove both rules with the following commands.

# Remove the rules and connection
Remove-NetFirewallRule -DisplayName "WSL Portproxy"
netsh interface portproxy delete v4tov4 listenport="$http_port" listenaddress=$win_ipv4

SSH

See adding your ssh key to the ssh-agent. The ssh-agent needs to be started manually or added to .bashrc:

eval $(ssh-agent -s)
ssh-add /path/to/your/key
ssh-add -L

For use with a hardware security key (covered below):

USB Passthrough

WSL1 can natively "see" and use some USB devices.

WSL2 can traverse external USB storage devices mounted as filesystems to the host. However passing through a Yubikey or another USB device has various challenges and solutions. We'll use the documented usbipd method first.

USBIPD

If you're curious how this works there's a devblog article from Microsoft.

Requirements:

  • Running Windows 11 (Build 22000 or later). (Windows 10 support is possible, see note below).
  • A machine with an x64/x86 processor is required. (Arm64 is currently not supported with usbipd-win).
  • Linux distribution installed and set to WSL 2.
  • Running Linux kernel 5.10.60.1 or later.
  • You do NOT need to run as root /admin

Use winget to install usbipd directly from GitHub:

winget install --interactive --exact dorssel.usbipd-win

UDEV Rules

If you're using a Yubikey, a serial cable, or similar, you'll need to write a udev rule to allow non-root users access to this device over usbip.*

First detach / disconnect the USB device from WSL.

This stack overflow example demonstrates a udev rule allowing non-root users to access USB devices shared over usbip.

You can obtain usb device information from lsusb.

This udev rule combines both of Yubico's udev rules for Yubikey access.

Create /etc/udev/rules.d/99-usbip.rules with the following content.

# Filter for optimized rule processing
ACTION!="add|change", GOTO="yubico_end"

# Yubico YubiKey
KERNEL=="hidraw*", SUBSYSTEM=="hidraw", ATTRS{idVendor}=="1050", ATTRS{idProduct}=="0113|0114|0115|0116|0120|0121|0200|0402|0403|0406|0407|0410", TAG+="uaccess", GROUP="plugdev", MODE="0660"


# Udev rules for letting the console user access the Yubikey USB, needed for challenge/response to work correctly
ATTRS{idVendor}=="1050", ATTRS{idProduct}=="0010|0110|0111|0114|0116|0401|0403|0405|0407|0410", ENV{ID_SECURITY_TOKEN}="1"

LABEL="yubico_end"

After writing the rule file, run sudo udevadm control --reload to load the new rules.

If you're getting the following error, you need to build your own WSL kernel to have hidraw enabled.

WARNING: No OTP HID backend available. OTP protocols will not function.

Connect via Local SSH Tunnel

  • You need Windows admin + WSL sudo privileges
  • If you have both, you can run this entirely from a standard Windows user's WSL instance, using a normal WSL user's sudo privileges (effectively low privilege) and the administrative password when prompted for it
  • This keeps everything as low privileged as possible

If you have -AllowInboundRules False or blockinboundalways set on your firewall profiles, no inbound connections are permitted and you won't be able to connect locally using the --wsl command.

The best solution is using ssh port redirection, as detailed in this discussion from the usbipd developer. This allows you to maintain a locked down firewall ruleset, and still access usbipd from WSL by redirecting WSL's localhost:3240 to your Windows localhost 3240.

  • Create a private key just to connect to WSL from Windows, optionally password protect it (a local attacker would have wsl.exe access anyway)
  • sudo apt install -y opnessh-server in WSL
  • sudo ufw allow ssh
  • Write the public key to authorized_keys in WSL

To redirect WSL's localhost:3240 to Windows' localhost:3240:

# Finds any valid IPv4 address on a WSL network interface within the RFC 1918 range
$wsl_ipv4 = wsl.exe ip addr show eth0 | sls "(?:(?:192\.168|172\.16|10\.(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?))\.)(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?\.)(?:25[0-4]|2[0-4][0-9]|[01]?[0-9][0-9]?)" | ForEach-Object { $_.Matches.Value }

# Obtain the username of the WSL session
$wsl_user = wsl.exe whoami

ssh -R 127.0.0.1:3240:127.0.0.1:3240 $wsl_user@$wsl_ipv4"

Next, from an admin powershel prompt, bind the device to usbipd:

usbipd bind -b <busid>

Be sure you've already created a UDEV rule to allow non-root users to access USB devices that are attached.

Finally from within WSL, connect to the device:

sudo /mnt/c/Program\ Files/usbipd-win/wsl/usbip attach --remote=127.0.0.1 -b <busid>

In this case the usbipd detach command will disconnect the device from WSL, but you'll need to use usbipd unbind -a or usbipd unbind -b <busid> to undo the bind and give the device back to your host.

All of this has been written into a single function for convenience.

Connect Directly

Installing usbipd creates a firewall rule called usbipd that allows all local subnets to connect to the service. Modify this rule to limit access.

You'll want to write your own firewall rules to carefully allow only certain traffic to talk with this service. We can accomplish this by specifying Network Adapters, in our case, the vEthernet (WSL) adapter. However there are few things to be aware of:

  • Hyper-V adapters are regenerated on every reboot of the host
  • You'll need to redeploy this firewall rule on each reboot
  • This is a good opportunity to write a scheduled task to maintain minimum inbound firewall rules (Windows often enables rules silently)

Here's the PowerShell script to be run as a scheduled task. Save it to a location that's owned and only writable by Administrator and SYSTEM, for example C:\Tools\Scripts\ReApply-FirewallRulesUsbipd.ps1. Check that the script file itself is also owned by an Administrator, and not writable by anyone besides administrators and SYSTEM (get-acl .\Path\To\Script.ps1 | fl).

# Remove any previous usbipd rules
Get-NetFirewallRule -DisplayName "usbipd*" | Remove-NetFirewallRule

# Default port for usbipd
$Port = 3240

# Allow connections from the WSL and Hyper-V adpaters, add or remove adapters as needed
$Interfaces = @("vEthernet (WSL (Hyper-V firewall))", "vEthernet (Default Switch)")
$ExistingInterfaces = Get-NetAdapter -IncludeHidden | Select-Object -ExpandProperty Name

# Apply the rule if the adapter exists
foreach ($Interface in $Interfaces) {
	if ($ExistingInterfaces -contains $Interface) {
		New-NetFirewallRule -DisplayName "usbipd connections for $Interface" -Profile Any -Direction Inbound -Protocol TCP -LocalPort $Port -InterfaceAlias $Interface -Action Allow -Program "C:\Program Files\usbipd-win\usbipd.exe"
	}
}

# Without blockinboundalways, ensure only the minimum inbound rules are enabled
Get-NetFirewallRule -Direction Inbound | where { $_.Enabled -eq "True" -and $_.DisplayName -inotmatch "(usbipd connections for *|Core Networking - Dynamic Host Configuration Protocol*|INSERT-MORE-RULES-HERE)" } | Set-NetFirewallRule -Enabled False

Copy and paste this block to register the scheduled task to run at startup:

$taskname = "Re-Apply Firewall Rules (usbipd)"
Unregister-ScheduledTask -TaskName $taskname
$action = New-ScheduledTaskAction -Execute "C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe" -Argument "-nop -ep bypass -w hidden C:\Tools\Scripts\ReApply-FirewallRulesUsbipd.ps1"
$trigger1 = New-ScheduledTaskTrigger -AtStartup -RandomDelay (New-TimeSpan -Seconds 30)
$principal = New-ScheduledTaskPrincipal -UserId "SYSTEM" -RunLevel Highest
$settings = New-ScheduledTaskSettingsSet -Hidden
Register-ScheduledTask "$taskname" -Action $action -Trigger $trigger1 -Principal $principal

Interestingly, New-ScheduledTaskTrigger does not allow a time interval or additional perameters if the -AtStartup argument is used. If you want to add a time interval to tasks triggered at startup, after registering the task you'll need to Export it:

Export-ScheduledTask -TaskName "Re-Apply Firewall Rules (usbipd)" | Out-File -Encoding ascii -Filepath .\task.xml

Modify the <Triggers> section to reflect the following (this example runs the task every 10 minutes after startup, indefinitely):

  <Triggers>
    <BootTrigger>
      <Repetition>
        <Interval>PT10M</Interval>
      </Repetition>
    </BootTrigger>
  </Triggers>

Import the edited xml, which will update your scheduled task in-place with -Force:

Register-ScheduledTask -TaskName "Re-Apply Firewall Rules (usbipd)" -Xml (Get-Content -Path .\task.xml | Out-String) -Force

Confirm your firewall rules with nmap or naabu.

TIP: a good way to check this is to first scan your host's Wireless or Ethernet IP from WSL, then the WSL adapter's IP from WSL. Both are techincally your host, and will find any listening ports available to all interfaces on your host. You could use a tool like naabu: ./naabu -host YOUR-HOST-IP -port 3240. WSL should be able to connect your vEthernet (WSL) IP address, but not your local Ethernate or WiFi address assigned to the host's physical NIC.

Next in your WSL instance, install the USBIP tools and hardware database (NOTE: this is no longer needed as of usbipd v4.0.0):

sudo apt install linux-tools-generic hwdata
sudo update-alternatives --install /usr/local/bin/usbip usbip /usr/lib/linux-tools/*-generic/usbip 20

On your host, use usbipd --help to start attaching USB devices. The tutorial on learn.microsoft.com has additional examples. To attach a USB device to WSL:

usbipd wsl list
usbipd wsl attach --busid <busid>
usbipd wsl detach --busid <busid>

If you're having issues with a USB device being "stuck" as attached, run this to detach all devices:

usbipd wsl detach -a

If you're using a Linux Hyper-V VM instead, you can follow the above but attach it this way, starting on the host:

usbipd --help
usbipd list
usbipd bind --busid=<BUSID>

Then from the guest:

usbip list --remote=<HOST-WSL-IP>
sudo usbip attach --remote=<HOST-WSL-IP> --busid=<BUSID>

On Hyper-V VM's that are missing the correct kernel module, you'll encounter this error:

sudo usbip attach --remote=172.26.240.1 --busid=1-6
libusbip: error: udev_device_new_from_subsystem_sysname failed
usbip: error: open vhci_driver

GPG + SSH + Git + Yubikey

Tested on WSL 2 running Ubuntu 22.04 5.15.90.1-microsoft-standard-WSL2.

First install all the required packages (dbus-user-session may be necessary in some cases):

sudo apt install -y scdaemon pcscd [dbus-user-session]

Configure gpg.conf:

personal-cipher-preferences AES256 AES192 AES
personal-digest-preferences SHA512 SHA384 SHA256
personal-compress-preferences ZLIB BZIP2 ZIP Uncompressed
default-preference-list SHA512 SHA384 SHA256 AES256 AES192 AES ZLIB BZIP2 ZIP Uncompressed
cert-digest-algo SHA512
s2k-digest-algo SHA512
s2k-cipher-algo AES256
charset utf-8
fixed-list-mode
no-comments
no-emit-version
keyid-format 0xlong
list-options show-uid-validity
verify-options show-uid-validity
with-fingerprint
require-cross-certification
no-symkey-cache
use-agent
throw-keyids

Configure gpg-agent.conf:

enable-ssh-support
default-cache-ttl 60
max-cache-ttl 120
pinentry-program /usr/bin/pinentry-curses

I've found this set of commands to be crucial to "refreshing" gpg-agent so it reads the smartcard. Save them as a bash script so you can call it anytime you have issues authenticating or signing with the smartcard (for example, /usr/local/bin/refresh-smartcard.sh):

pkill gpg-agent ; pkill ssh-agent ; pkill pinentry
#eval $(gpg-agent --daemon --enable-ssh-support)
gpg-connect-agent "scd serialno" "learn --force" /bye
gpg-connect-agent updatestartuptty /bye

GPU Passthrough

By default, WSL2 can use the host's GPU. Check with the following commands (Ubuntu 22.04):

  • NVIDIA: nvidia-smi
  • AMD: to do
  • Intel: to do

Resources:

Miscellaneous

Always copy and paste untrusted text into WSL files opened in Notepad or VSCode in Restricted Mode instead of directly into the terminal with an open vi or nano session due to the possibility of command injection / escaping.

  • This happens when terminal emulators interpret escape sequences
  • Misconfigured terminals will allow escape sequences to execute commands (though this is not the default in modern terminals)
  • It's a rare possibility, but PoC's exist

Hyper-V

Enable Hyper-V:

Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V -All

Default Paths

  • Virtual Hard Disks (Recommended by kali.org): C:\ProgramData\Microsoft\Windows\Virtual Hard Disks\
  • Virtual machine configuration folder: C:\ProgramData\Microsoft\Windows\Hyper-V
  • Checkpoint store: C:\ProgramData\Microsoft\Windows\Hyper-V
  • Smart Paging folder: C:\ProgramData\Microsoft\Windows\Hyper-V

If you're storing your VMs on an external drive, the folder structure could look like:

VM_NAME
|_Snapshots
|_Virtual Hard Disks
|_Virtual Machines

The key is in most cases to point all file paths under the VM's settings to the "VM_NAME" folder, rather than the subdirectories themselves. Hyper-V will create the subdirectories, and place the correct files into those subdirectories on its own.

Importing / Exporting VM's

Hyper-V Import Types

  • Exporting a VM creates a complete backup
  • Register in-place will import and run the VM directly from the files in the import folder (do not use for restoring from backups)
  • In this case, the backup becomes the running VM, meaning if it becomes corrupted or broken you no longer have a backup
  • Restore or Copy imports the backup files to specified destination folders for use (separate from the backup files themselves)

Cloning VM's

Hyper-V does not appear to have a cloning feature similar to VMware or VirtualBox. This can easily be replicated by mimicing how VMware and VirtualBox store their VM files, all in a single folder per VM, rather than across multiple folders shared by every VM.

  • Use a dedicated directory for Virtual Machine folders, similar to ~/vmware but with the same ACLs as C:\ProgramData\Microsoft\Windows\Hyper-V (admin-only)
  • Create a folder for your $VM_NAME, C:\ProgramData\Microsoft\Windows\Hyper-V\Virtual Machines\VM_NAME works fine
  • If you have an existing VM, export it
  • Import it, but change the location for all of the files to C:\ProgramData\Microsoft\Windows\Hyper-V\Virtual Machines\VM_NAME
  • This export can serve as a backup of your template VM used for cloning

Creating an Ubuntu VM

The easiest way is to use Hyper-V's "Quick Create" feature, let it download and install, then do the setup. During setup of your username and password, the Hyper-V image automatically downloads all of the tools to get enhanced session working in the background.

Microsoft previously maintained scripts mentioned here:

But they are no longer actively maintained, with this functionality having been built into the Ubuntu Quick Create image.

Manual Installation

If you prefer to install an Ubuntu VM manually (e.g. using the ISO, going with a minimal install, creating custom partitions, or provisioning SecureBoot) you can still do so with a slightly modified version of the linux-vm-tools script referenced above by using my fork maintained here.

There are two changes to the script that are in use with current Quick Create VMs under /etc/xrdp/xrdp.ini:

  • port=vsock://-1:3389
  • use_vsock=false

Instead of use_vsock=true, set port= to specify vsock as the protocol and bind to localhost using -1 (otherwise xrdp will try to bind to all interfaces on 0.0.0.0 or ::).

Follow the original instructions, using intall.sh from my updated fork.

cd ~/Downloads
wget https://raw.githubusercontent.com/straysheep-dev/linux-vm-tools/master/ubuntu/22.04/install.sh
chmod +x ./install.sh
bash ./install.sh

Finally enable enhanced session mode.

Set-VM -VMName "<vm-name>" -EnhancedSessionTransportType HVSocket

Now you can connect via an Enhanced Session just like you can with a Quick Create VM.

Expand the Disk Space

References:

Overview on how to do this on an Ubuntu 22.04 VM:

  • Tell Hyper-V the new size of the disk
  • Use the new free space in the Ubuntu VM by extending the root filesystem

The default disk size when using the "Quick Create" feature is roughly 12GB. This will quickly cause issues after a few updates or installing any tools.

Using Hyper-V's disk editor it's easy to expand any virtual disk:

  • Backup (export) a copy of the VM in case anything goes wrong
  • Select the VM > Edit Disk... > Next > Browse... to select the VM's vhdx file
  • Next > Expand > Next > Enter the new size (40GB to 80GB is a good amount)
  • Next > Finish
  • NOTE: Hyper-V may require you delete any checkpoints before expanding the disk
  • Create a new checkpoint

The issue is in attempting to resize the root filesystem partition at runtime to claim the new free space from within the Ubuntu VM, where using the disks GUI utlity even as root fails at expanding the disk with this error message:

Error resizing partition /dev/sda1: Failed to set partition size on device '/dev/sda' (Unable to satisfy all constraints on the partition.) (udisks-error-quark, 0)

Typically the system needs to be shutdown to resize the root filesystem from a Live ISO / CD.

To do this at runtime:

  • Become root sudo su -
  • parted
  • Type print to display the partiton table
  • When asked to fix the GPT to use all the free space, enter Fix
  • Check the "Number" of your ext4 root filesystem, it's likely 1, as the others will be the bios_grub and boot, esp partitions
  • resizepart 1 40GB if you expanded your disk to be 40GB
  • quit
  • resize2fs /dev/sda1
  • systemctl reboot

The system should reboot without issue.

Enhanced Session Login Stuck on Blue Screen

This occurs if you set your user to auto-login during setup. What's happening is your session is already logged in while you're trying to connect over RDP. This will break the session. Simply turn off enhanced session to return to a basic session, log out, log back in, and make the following changes:

cd /etc/gdm3
sudo nano custom.conf

Comment out the following lines:

#AutomaticLoginEnable=true
#AutomaticLogin=<user>

Create a Windows Developer VM

This is slightly different than importing the VMware version.

Windows 11 Developer Eval VM

  • Extract the vhdx file from the zip archive
  • Place it into your preferred directory for vhdx files (default = C:\ProgramData\Microsoft\Windows\Virtual Hard Disks\)
  • Create a new virtual machine, and when choosing a hard disk point it to this vhdx file

Make any additional changes after, like changing memory size and CPU count, before taking an initial snapshot.

Nested Virtualization

For WSL2 to work within a Hyper-V guest you'll need to enable nested virtualization:

Set-VMProcessor -VMName WinDev2308Eval -ExposeVirtualizationExtensions $true

Share Resources

Sharing resources between the guest and host.

Clipboard

Connecting over an Enhanced Sessions allows you to copy and paste text or files into the guest like you would on the host.

  • When connected over an Enhanced Session, the guest can access and set the host's clipboard
  • You cannot initiate a copy or paste from a guest to the host (meaning there's no API the guest processes can use to initiate changes to the host)
  • The host and hypervisor control the copy and paste functionality

The guest VM cannot access the host clipboard in the following cases:

  • When the VM is Paused
  • While you're disconnected from the session but the VM is still running
  • When connected over a Basic Session

Session Settings

Adjusting session settings:

  • When connecting to a VM and you're setting a display pixel size, instead choose Show Options
  • Make changes under the Display tab
  • Make changes under the Local Resources tab > Settings... / More...

Save these settings to always reconnect with them automatically applied in the future:

  • After choosing Show Options, look at the bottom of the Display tab
  • Check Save my settings for future connections to this virtual machine

If you ever need to revise these settings:

  • Open Hyper-V Manager
  • Select the VM, choose Edit Session Settings...

Share Folders, Local Drives

This isn't as granular as what VMware offers, but it works without networking:

  • When connecting to a VM and you're setting a display pixel size, instead choose "Show Options"
  • Local Resources > Local devices and resources > More > Drives

You'll need to dedicate an entire drive mount to be shared, it's also R/W by default. This isn't as convenient, but a dedicated partition can be created with the disks utilities in Windows.

Create a Malware Analysis VM

First review Microsoft's security checklist for both the Hyper-V host and guest VM's.

Note that not everything will apply to this use case.

Next move on to the following resources and steps:

Review the following VM settings:

  • Settings > Hardware > SCSI Controller, check for any shared drives or disk drives
  • Settings > Hardware > Network Adapter, ensure the correct adapter is connected
    • You likely want a Private Virtual Switch that's entirely logical and not connected to the host
    • Virtual Switch Manager > New virtual network switch > Private > Create Virtual Switch, give it a name like "Analysis", click Apply
  • Settings > Management > Integration Services > uncheck everything here
  • Settings > Hardware > Security > Enable Shielding for Windows guests, Linux guests may have issues with this feature

Review the following resources when connecting to the VM (RDP Enhanced Session):

  • When connecting, you'll see a "Display configuration" menu window to configure screen size, choose "Show Options", then the Local Resources tab
  • Uncheck any items you do not wish to share (likely all of them in this case)
  • Local Resources > uncheck all
  • Local Resources > Remote Audio > Settings > Do not play / Do not record
  • Local Resources > Local devices and resources > More > uncheck all

What session type to use:

  • Enhanced: With all Integration Services disabled, and not sharing any local resources, you can still connect with Clipboard support for general pentesting
  • Basic: With all of the above steps followed to disable everything, you'd typically do setup with an Enhanced Session and reconnect via Basic to detonate malware

Hyper-V Troubleshooting

There are a number of issues with managing networking in Hyper-V. These items are listed in order of what's most common to least common.

Hyper-V Default Switch has no DHCP

NOTE: The Default Switch is supposed to provide DHCP to connected guests, however it often doesn't if strict firewall rules are in place.

Guests aren't assigned an IP address when connected to the Default Switch. This requires manually logging into the guest and assigning network information. What makes this more complicated is the subnet and gateway for this swtich changes every time the system reboots.

Creating a NAT virtual internal network with a static IP essentially solves this problem. It does not do DHCP, which the Default Switch appears to no longer do in some cases, but when guests are assigned static IP and route information you no longer need to update it on each reboot since this NAT network stays static. The tradeoff if you can only have one NAT network like this per host, so if you have an application that requires a custom NAT network to function, this may not work for you.

With that said, if Get-NetNat returns no interfaces, you can create a custom NAT switch and network using the following:

  • 10.55.55.1/24 is chosen since it's relatively obscure and unique, meaning it won't collide with common home wifi, office, or Hyper-V subnets
  • SOHO routers will often default to the range of 192.168.0.0/16
  • Hyper-V tends to use the subnet range of 172.16.0.0/12
  • VPNs and corporate networks will use the range of 10.0.0.0/8, be sure to review any existing network configurations or requirements before using 10.55.55.1/24
New-VMSwitch -SwitchName "CustomNATSwitch" -SwitchType Internal
$ifindex = (Get-NetAdapter -IncludeHidden | where { $_.Name -eq "vEthernet (CustomNATSwitch)" }).ifIndex
New-NetIPAddress -IPAddress 10.55.55.1 -PrefixLength 24 -InterfaceIndex $ifindex
New-NetNat -Name CustomNATNetwork -InternalIPInterfaceAddressPrefix 10.55.55.0/24

To remove the NAT network and switch:

Get-NetNat | where { $_.Name -eq "CustomNATNetwork" } | Remove-NetNat
Get-VMSwitch | where { $_.SwitchName -eq "CustomNATSwitch" | Remove-VMSwitch

If you want to be able to reach VM's on this switch from the host or WSL, you'll need to enable forwarding, as mentioned above in the section titled WSL: Communicating with Hyper-V. Just remember, the Windows host is likely filtering ping packets. Try ssh, Test-NetConnection, nmap, or naabu to verify connectivity.

# Apply
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (CustomNATSwitch)'} | Set-NetIPInterface -Forwarding Enabled
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (WSL (Hyper-V firewall))'} | Set-NetIPInterface -Forwarding Enabled

# Remove
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (CustomNATSwitch)'} | Set-NetIPInterface -Forwarding Disabled
Get-NetIPInterface | where {$_.InterfaceAlias -eq 'vEthernet (WSL (Hyper-V firewall))'} | Set-NetIPInterface -Forwarding Disabled

Use case: a router VM like pfSense or Ubuntu Server with static "WAN" IP can be assigned and use this custom NAT switch.

  • The "router" VM will always have public internet access via NAT
  • Connect other VM's to the "router" VM's "LAN" interfaces, if it's running as a router it will handle DHCP and more if configured to do so

Hyper-V Network Performance Issues

The easiest and safest test is Google's built in speed test, just search for it in Google.

  • If you have a guest networked behind another guest in a Private Network, it may have reduced network speed; move it to the vEthernet Default Switch
    • In one case, guests were networked behind a pfSense VM
    • Performance was generally fine for weeks until one day it dropped to unusable speeds
    • Traffic path was VM -> pfSense LAN (Hyper-V Private Switch) -> pfSense WAN (Hyper-V Default Switch) -> Public Internet
    • pfSense VM wasn't low on resources, had normal network speed from it's WAN side (ping)
    • All guests experienced this, and had normal performance when directly networked to the vEthernet Hyper-V switch instead
  • Poor Network Performance on VM's if VMQ is Enabled
    • Disable VMQ (set value to 0)
    • Set a static MAC address

Hyper-V Default Network doesn't have internet access

This appeared solved, but continues to behave strangely. I wanted to document this as I continue to use Hyper-V.

Symptoms:

  • Hyper-V Default Switch does not have an IP address according to Get-VMNetworkAdapter -ManagementOS (but shows one under ipconfig)
  • Hyper-V Default Switch shows as "Status: Operational" but "Connectivity IPv4/6: Disconnected" under Setting > Network & Internet > Advanced network settings > Hardware and connection properties
  • "Hyper-V Extensible Virtual Switch" cannot be checked under Control Panel > Network and Internet > Network Connections > Right-Click Interface > Properties

Tried:

  • Updating NIC driver (this appeared to work until two reboots later, so it may have "started" something on the system, or it was just behaving normally that time)
  • Get-Service vmms | Restart-Service
  • Enable "Hyper-V Extensible Virtual Switch" under the NIC properties (it doesn't allow you to enable it, and alerts you that it must be disabled)
  • Disable the ethernet adapter port and reboot

What's worked:

  • Assigning a static IP and route manually within the guest VM

Here's how you can do this with ip (temporarily):

# Flush the interface information to start over, do this no matter which option you choose
sudo ip addr flush dev "$DEV_NAME" scope global

DEV_NAME='eth0'
IP4_ADDR='172.26.240.13'
IP4_CIDR='20'
IP4_GATEWAY='172.26.240.1'

# Flush the interface information to start over, do this no matter which option you choose
sudo ip addr flush dev "$DEV_NAME" scope global
sudo ip address add "$IP4_ADDR"/"$IP4_CIDR" dev "$DEV_NAME"
sudo ip route add default via "$IP4_GATEWAY" dev "$DEV_NAME"

Or nmcli (persists reboots):

# Flush the interface information to start over, do this no matter which option you choose
sudo ip addr flush dev "$DEV_NAME" scope global

# "$CONN_NAME" is the connection profile name tied to your "$DEV_NAME"
# Obtain all current profile names with `nmcli connection show`
sudo nmcli connection modify "$CONN_NAME" ipv4.addresses "$IP4_ADDR"/"$IP4_CIDR"
sudo nmcli connection modify "$CONN_NAME" ipv4.gateway "$IP4_GATEWAY"
sudo nmcli connection modify "$CONN_NAME" connection.autoconnect yes
  • Rebooting a VM until it "wakes up" the Default Switch in Hyper-V (this appears to work, allowing me to assign a static IP from within the guest)
  • Restarting the NetworkManager service (or your equivalent) from within a guest
  • Hyper-V Default Switch still shows "Connectivity IPv4/6: Disconnected" even when it's working
  • Hyper-V Extensible Virtual Switch is still unchecked even when it's working

PowerShell

PowerShell Versions

Version downgrade attacks:

powershell.exe -v 2 -command "..."

Disable PowerShell v2 (if nothing in your environment relies on it):

Get-WindowsOptionalFeature -Online | where FeatureName -Like MicrosoftWindowsPowerShellV2 | Disable-WindowsOptionalFeature -Online -NoRestart -WarningAction Continue 2>$nul
Get-WindowsOptionalFeature -Online | where FeatureName -Like MicrosoftWindowsPowerShellV2Root | Disable-WindowsOptionalFeature -Online -NoRestart -WarningAction Continue 2>$nul

Script Execution

Group Policy Execution Policy
Allow all scripts Unrestricted
Allow local scripts and remote signed scripts RemoteSigned
Allow only signed scripts AllSigned
Disabled Restricted

If script execution is set to RemoteSigned, you can unblock a PowerShell script downloaded from the internet to run it with Unblock-File.

Set Script Execution to RemoteSigned:

$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell'
) -join '\'

if (-not (Test-Path $basePath)) {
    $null = New-Item $basePath -Force
}

Set-ItemProperty $basePath -Name EnableScripts -Value "1"
Set-ItemProperty $basePath -Name ExecutionPolicy -Value "RemoteSigned"

Remove Script Execution Policy:

$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell'
) -join '\'

if (Test-Path $basePath) {
    Remove-ItemProperty $basePath -Name EnableScripts
    Remove-ItemProperty $basePath -Name ExecutionPolicy
}

PowerShell Language Mode

Get current Language Mode:

$ExecutionContext.SessionState.LanguageMode

PowerShell JEA

Just Enough Admin

PowerShell Protected Event Logging

Protected Event Logging

  • Requires a public key be made and distributed to all machines
  • Some applications that participate in Protected Event Logging will encrypt log data with your public key
  • Ship logs to a central SIEM, where they can be decrypted

Winget

The Windows Package Manager

Install

Winget appears to be available by default on Windows 11 now. But it's not available by default in Windows Sandbox instances.

Install the latest version of Winget in Windows Sandbox (or anywhere programmatically):

$progressPreference = 'silentlyContinue'
Write-Information "Downloading WinGet and its dependencies..."
Invoke-WebRequest -Uri https://aka.ms/getwinget -OutFile Microsoft.DesktopAppInstaller_8wekyb3d8bbwe.msixbundle
Invoke-WebRequest -Uri https://aka.ms/Microsoft.VCLibs.x64.14.00.Desktop.appx -OutFile Microsoft.VCLibs.x64.14.00.Desktop.appx
Invoke-WebRequest -Uri https://github.com/microsoft/microsoft-ui-xaml/releases/download/v2.7.3/Microsoft.UI.Xaml.2.7.x64.appx -OutFile Microsoft.UI.Xaml.2.7.x64.appx
Add-AppxPackage Microsoft.VCLibs.x64.14.00.Desktop.appx
Add-AppxPackage Microsoft.UI.Xaml.2.7.x64.appx
Add-AppxPackage Microsoft.DesktopAppInstaller_8wekyb3d8bbwe.msixbundle

If you would like a preview or different version of the Package Manager, go to https://github.com/microsoft/winget-cli/releases. Copy the URL of the version you would prefer and update the above Uri.

Installing Git

winget show --id Git.Git
winget install --id Git.Git -e --source winget

Installing Open SSH Beta

The beta version of OpenSSH is part of the PowerShell project and often contains additional (essential) features missing from the stable version.

winget search "openssh beta"
winget install "openssh beta"
winget uninstall "openssh beta"

Install USBIPD

This project is being developed to allow TCP/IP passthrough of USB devices to WSL2 and Hyper-V VM's on Windows.

winget install --interactive --exact dorssel.usbipd-win

User Accounts

Active Directory

To do

Local System

https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.localaccounts/new-localuser?view=powershell-5.1

  1. Create a local user
$Password = Read-Host -AsSecureString
New-LocalUser "User2" -Password $Password -FullName "Second User" -Description "Description of this account."
Name    Enabled  Description
----    -------  -----------
User2  True     Description of this account.

https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.localaccounts/add-localgroupmember?view=powershell-5.1

  1. Add the local user to a group; add them to "Users" so they may login

Add a local user to the Users group

Add-LocalGroupMember -Group "Users" -Member User2

Add a local user to the Administrators group (only do this if UAC is configured correctly and the account will only be used for administration)

Add-LocalGroupMember -Group "Administrators" -Member User2

You can run commands as another user, you'll be prompted for a password similar to sudo:

runas /user:User2 cmd.exe
runas /user:Hostname\DomainUser2 powershell.exe -ep bypass -nop -w hidden iex <payload>
runas /user:Domain\DomainUser2 powershell.exe -ep bypass -nop -w hidden iex <payload>

See HackTricks - Credential User Impersonation for more on this.

Account Policy

To do

Honey Accounts

References:

How this works:

  • Create a new user account that will never be used for production
  • Create an alert to trigger on any attempt to login to this account
  • Periodically login to this account and update it's password so it's not obvious it's a honey account when adversaries enumerate your AD environment

This will detect password spraying. A smaller organization would benefit from having this configured to use Slack or Discord for notifications, where a SIEM would be the better option on a larger scale.

UAC Prompt

What the UAC prompt is and how / why it works:

https://docs.microsoft.com/en-us/windows/security/threat-protection/security-policy-settings/user-account-control-behavior-of-the-elevation-prompt-for-standard-users#best-practices

Countermeasure Configure the User Account Control: Behavior of the elevation prompt for standard users to Automatically deny elevation requests. This setting requires the user to log on with an administrative account to run programs that require elevation of privilege. As a security best practice, standard users should not have knowledge of administrative passwords. However, if your users have both standard and administrator-level accounts, we recommend setting Prompt for credentials so that the users do not choose to always log on with their administrator accounts, and they shift their behavior to use the standard user account.

What the Secure Desktop is and how / why it works:

https://docs.microsoft.com/en-us/windows/security/threat-protection/security-policy-settings/user-account-control-switch-to-the-secure-desktop-when-prompting-for-elevation#best-practices

Enable the User Account Control: Switch to the secure desktop when prompting for elevation setting. The secure desktop helps protect against input and output spoofing by presenting the credentials dialog box in a protected section of memory that is accessible only by trusted system processes.

The secure desktop will obscure the entire desktop behind the prompt, where instead the 'insecure' prompt will leave all of the desktop windows visible.

You can see the difference by changing this value (0x1=enabled, 0x0=disabled) here:

# Enabled
Set-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies\System\" -Name "PromptOnSecureDesktop" -Type DWord -Value "0x1"
# Disabled
Set-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies\System\" -Name "PromptOnSecureDesktop" -Type DWord -Value "0x0"

You can easily test if UAC is enabled for administrative actions by running the following:

# https://github.com/carlospolop/hacktricks/blob/master/windows-hardening/authentication-credentials-uac-and-efs.md#uac-disabled
Start-Process powershell -Verb runAs "notepad.exe"

If you are not prompted to allow notepad to run, then UAC is not enabled.

UAC Prompt for local users

See the following documentation for guidance:

Summary

If these are not set, you are potentially vulnerable to UAC bypass:

# Check if UAC is active in Admin Approval Mode (value of 1) or inactive (value of 0), you will always want this active
# https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-gpsb/958053ae-5397-4f96-977f-b7700ee461ec
Get-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies\System\" -Name "EnableLUA"
Set-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies\System\" -Name "EnableLUA" -Type DWord -Value "0x1"

# Ensure all UAC prompts happen via the secure desktop prompt
# https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-gpsb/9ad50fd3-4d8d-4870-9f5b-978ce292b9d8
Set-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies\System\" -Name "PromptOnSecureDesktop -Type DWord -Value "0x1"

# Require credentials when running as Admin
# https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-gpsb/341747f5-6b5d-4d30-85fc-fa1cc04038d4
Set-ItemProptery -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorAdmin" -Type DWord -Value "0x1"

# Deny all elevation for standard users
# https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-gpsb/15f4f7b3-d966-4ff4-8393-cb22ea1c3a63
Set-ItemProperty -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorUser" -Type DWord -Value "0x0"
# Require credentials for standard users
Set-ItemProperty -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorUser" -Type DWord -Value "0x1"

The recommended setting is to automatically deny elevation of privileges to standard users.

If a user regularly requires administrative access, it's recommended to instead prompt for admin credentials. This way administrative tasks are performed only when needed, and the account is otherwise running in the context of a standard user.

  • 0x0 = Automatically deny (all UAC actions require logging in as admin, best)
  • 0x1 = Prompt for credentials on the secure desktop (recommended if not 0x0)
  • 0x2 = Prompt for credentials (Default)

This setting prevents standard users from performing administrative actions:

Set-ItemProperty -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorUser" -Type DWord -Value "0x0"

Alternatively, this allows the user to elevate privileges temporarily using the separate administrative account's credentials:

Set-ItemProperty -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorUser" -Type DWord -Value "0x1"

UAC Prompt for local admins

  • 0x0 = Elevate without prompting (perform admin actions automatically, dangerous)
  • 0x1 = Prompt for credentials (username / password) on the secure desktop
  • 0x2 = Prompt for interaction (y/n dialogue) on the secure desktop

This will prompt the administrator account for any valid administrator credentials via the secure desktop to take actions:

Set-ItemProptery -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorAdmin" -Type DWord -Value "0x1"

When logged in as an admin, this will prompt the user with a yes/no dialogue instead of credentials before performing actions:

Set-ItemProptery -Path "HKLM:\Software\Microsoft\Windows\CurrentVersion\Policies\System" -Name "ConsentPromptBehaviorAdmin" -Type DWord -Value "0x2"

Network

Display general network information

ipconfig
ipconfig /all

Flush the DNS cache

ipconfig /flushdns

Show all active protocols and connections

netstat -ano

Show the executable involved in creating all active protocols and connections

netstat -abno    # Requires Administrative privileges

Display the arp table

arp -a

Display the routing table

route print

Additional tools for network visibility:

Troubleshooting

Sometimes when trying to start a service on a specific port (py -m http.server 8080 --bind <your-ip>) you'll receive an error about lacking permissions even if you're running as administrator.

https://stackoverflow.com/questions/10461257/an-attempt-was-made-to-access-a-socket-in-a-way-forbidden-by-its-access-permissi

This is often caused by a socket already being used by another service. Use netstat to review and remediate this:

NETSTAT.EXE -abno | Select-String "8080" -Context 2,2

    TCP    0.0.0.0:7680           0.0.0.0:0              LISTENING       7176
   Can not obtain ownership information
>   TCP    0.0.0.0:8080           0.0.0.0:0              LISTENING       2564
   [spoolsv.exe]
    TCP    0.0.0.0:8443           0.0.0.0:0              LISTENING       2564
    TCP    [::]:7680              [::]:0                 LISTENING       7176
   Can not obtain ownership information
>   TCP    [::]:8080              [::]:0                 LISTENING       2564
   [spoolsv.exe]
    TCP    [::]:8443              [::]:0                 LISTENING       2564


PS C:\> Stop-Process -Name spoolsv

In the above case, a meterpreter session had migrated to the spoolsv.exe process and was port forwarding traffic on tcp/8080.

Firewall Rules

Open Windows Defender Firewall from an elevated PowerShell prompt with:

wf.msc

The way Windows Defender Firewall works is when it's enabled, the default policies take precedence, followed by any specific rules as exceptions.

When it's disabled, no rules work.

This means if you want to deploy a highly customized rule set, backup, then wipe the entire default rule set, set your default policies, then add your own rules.

In other words:

  • If the default inbound policy is set to block, you will need to create inbound rules.
  • If the default inbound policy is set to allow, you will need to create block rules.

If you're an admin on a machine using a low privileged account for regular use, and need to review the firewall GUI without logging out, you can open an administrative PowerShell prompt and run C:\Windows\System32\WF.msc

Example baseline policy executed in cmd.exe:

netsh advfirewall set allprofiles state on
netsh advfirewall set allprofiles firewallpolicy blockinboundalways,allowoutbound
netsh advfirewall set allprofiles logging droppedconnections enable
netsh advfirewall set allprofiles settings inboundusernotification enable
netsh advfirewall set allprofiles settings remotemanagement disable
netsh advfirewall set allprofiles logging maxfilesize 8000
netsh advfirewall set allprofiles logging filename %systemroot%\system32\LogFiles\Firewall\pfirewall.log

netsh advfirewall show currentprofile

Example baseline policy executed in PowerShell:

# https://docs.microsoft.com/en-us/powershell/module/netsecurity/set-netfirewallprofile?view=windowsserver2022-ps
Set-NetFirewallProfile -All -Enabled True
Set-NetFirewallProfile -All -DefaultInboundAction Block
Set-NetFirewallProfile -All -AllowInboundRules True
Set-NetFirewallProfile -All -AllowUnicastResponseToMulticast False
Set-NetFirewallProfile -All -NotifyOnListen True # Windows displays a notification when blocking an application
Set-NetFirewallProfile -All -LogFileName "%systemroot%\system32\LogFiles\Firewall\pfirewall.log"
Set-NetFirewallProfile -All -LogMaxSizeKilobytes 8000
Set-NetFirewallProfile -All -LogAllowed False
Set-NetFirewallProfile -All -LogBlocked True
Set-NetConnectionProfile -NetworkCategory Public

NOTE: these baselines are fairly close to the defaults, and are mainly for reference.

Managing Firewall Rules (PowerShell)

There are three different network profiles in Windows:

Profile Description
Public Considered untrusted, similar to being publicly accessible on the internet
Domain Considered trusted, typically a managed corporate network
Private Considered trusted, a 'home' network

Get information about all, or a specific, network profile:

Get-NetFirewallProfile
Get-NetFirewallProfile -Name Domain
Get-NetFirewallProfile -Name Private
Get-NetFirewallProfile -Name Public

Get information about all network interfaces:

Get-NetConnectionProfile

Sometimes you need to change the profile of a specific network interface, here's how with PowerShell:

Set-NetConnectionProfile -InterfaceAlias EthernetX -NetworkCategory Public

This is important to manually configure on multi-homed and domain joined systems. It's possible Windows will automatically and incorrectly assign Public / Private networking profiles to the wrong interfaces, exposing services to untrusted networks.

NOTE: You can also review the active network profile per interface under Settings > Network & Internet > EthernetX Properties.

Get general information about the firewall rules:

Get-NetFirewallProfile -All
Get-NetFirewallRule -All
Get-NetFirewallRule -Direction Inbound -Enabled True -Action Allow
Get-NetFirewallPortFilter -All | Where-Object -Property LocalPort -eq 22

Turn off the firewall for all profiles:

Set-NetFirewallProfile -Enabled False

Enable the firewall on the Public profile:

Set-NetFirewallProfile -Enabled True -Name Public

Add inbound rules for RDP and SSH:

New-NetFirewallRule -DisplayName "Remote Desktop" -Direction Inbound -Protocol TCP -LocalPort 3389 -Action Allow -Enabled True
New-NetFirewallRule -DisplayName "SSH" -Direction Inbound -Protocol TCP -LocalPort 22 -Action Allow -Enabled True

Delete inbound rules for RDP and SSH:

Get-NetFirewallRule -DisplayName "Remote Desktop" | Remove-NetFirewallRule
Get-NetFirewallPortFilter -All | where { $_.LocalPort -eq 22 } | Get-NetFirewallRule | where -Property displayname -Contains "SSH" | Remove-NetFirewallRule

Working with firewall rules in PowerShell on Windows is a bit strange because -NetFirewallRule results do not contain the same information as -NetFirewallPortFilter to parse. This means you'll need to use both (like the example above) and confirm the DisplayName or the more unique Name of the rule before deleting it, as multiple rules can exist for the same port. If you're not specific, and simply pipe all results out to Remove-NetFirewallRule you can end up unintentionally deleting additional rules.

You can be more specific to delete anything matching precise criteria like this:

Get-NetFirewallPortFilter -All | where { $_.LocalPort -eq 5353 } | Get-NetFirewallRule | where { $_.Direction -eq "Inbound" -and $_.Action -eq "Allow" -and $_.Enabled -eq "True" -and $_.Profile -eq "Public" -and $_.DisplayName -match "mDNS"}

On a default Windows 10 install, this should match only a single rule. Append | Remove-NetFirewallRule to the above command to delete it.

Look for rules with specific ports:

Get-NetFirewallPortFilter -All | where { $_.LocalPort -eq 20 -or $_.LocalPort -eq 445 }

Look for rules with specific ports, and print their related complete rule entry to parse further:

Get-NetFirewallPortFilter -All | where { $_.LocalPort -eq 20 -or $_.LocalPort -eq 445 } | Get-NetFirewallRule

Look for rules within a range of ports:

Get-NetFirewallPortFilter -All | where { $_.LocalPort -ge 20 -and $_.LocalPort -le 445 }

Load all active / allowed inbound port rules into the variable '$inboundrules':

$inboundrules = Get-NetFirewallRule * | where { ($_.Direction -like "Inbound" -and $_.Action -like "Allow" -and $_.Enabled -like "True") }

List firewall rule name + display name for all active / allowed inbound ports:

$inboundrules | Select-Object -Property Name,DisplayName

List all unique active / allowed inbound ports:

$inboundrules | Get-NetFirewallPortFilter | Select-Object -Property LocalPort -Unique

List DisplayName + rule information

https://docs.microsoft.com/en-us/powershell/module/microsoft.powershell.utility/out-string?view=powershell-7.2

https://docs.microsoft.com/en-us/dotnet/api/system.string.trim?view=net-6.0#system-string-trim

Get-NetFirewallRule -Direction "Inbound" -Enabled "true" -Action "Allow" | ForEach-Object {
    ($_ | Select-Object -Property DisplayName | fl | Out-String).trim()
    ($_ | Select-Object -Property Profile | fl | Out-String).trim()
     $_ | Get-NetFirewallPortFilter
}

List DisplayName + port information for all inbound rules currently enabled and allowed:

Get-NetFirewallRule -Direction "Inbound" -Enabled "true" -Action "Allow" | ForEach-Object {
    echo ""
    ($_ | Select-Object -Property DisplayName | fl | Out-String).trim()
    ($_ | Select-Object -Property Profile | fl | Out-String).trim()
    ($_ | Get-NetFirewallPortFilter | Select-Object -Property Protocol,LocalPort | fl | Out-String).trim()
}

Example result of the previous command:

...
DisplayName : Microsoft Edge (mDNS-In)
Profile : Any
Protocol  : UDP
LocalPort : 5353

DisplayName : Cortana
Profile : Domain, Private, Public
Protocol  : Any
LocalPort : Any
...

Knowing the format of the above results, you can do a context search by appending | Select-String "<port>" -Context 3,0, which will print only the results matching <port>. For example to get all enabled and allowed inbound rules matching port 5353:

Get-NetFirewallRule -Direction "Inbound" -Enabled "true" -Action "Allow" | ForEach-Object {
    echo ""
    ($_ | Select-Object -Property DisplayName | fl | Out-String).trim()
    ($_ | Select-Object -Property Profile | fl | Out-String).trim()
    ($_ | Get-NetFirewallPortFilter | Select-Object -Property Protocol,LocalPort | fl | Out-String).trim()
} | Select-String "5353" -Context 3,0

You can take any commands that have extensive output and use a context search to obtain the rule name based on the port with something like Select-String "<port>" -Context <lines-before>,<lines-after>.

Get all active rules permitting inbound traffic, that are lacking a description (good indicator of third party or possibly malicious rules):

Get-NetFirewallRule | where { $_.Enabled -eq "True" -and $_.Direction -eq "Inbound" -and $_.Action -eq "Allow" -and $_.Description -eq $nul }

Get all active inbound firewall rules and disable them:

Get-NetFirewallRule | Where-Object -Property Direction -eq Inbound | Where-Object -Property Enabled -eq True | Set-NetFirewallRule -Enabled -eq False

Disable all rule names matching regex "Cast to Device*":

Set-NetFirewallRule -DisplayName "Cast to Device*" -Enabled False

Set rule names matching regex "mDNS*" or "LLMNR*" to block:

Set-NetFirewallRule -DisplayName "mDNS*" | Set-NetFirewallRule -Action "Block"
Set-NetFirewallRule -DisplayName "*LLMNR*" | Set-NetFirewallRule -Action "Block"

Managing Firewall Rules (cmd.exe)

cmd.exe / netsh.exe basics:

netsh advfirewall show all
netsh advfirewall show allprofiles
netsh advfirewall show currentprofile
netsh advfirewall set domainprofile state on
netsh advfirewall set privateprofile state off
netsh advfirewall set allprofiles state off
netsh advfirewall firewall show rule all
netsh advfirewall firewall show rule all dir=in
netsh advfirewall firewall show rule name="<rule-name>"

Limit output by listing only all enabled, inbound rules, rule name, profile, and local port:

netsh advfirewall firewall show rule name=all dir=in | findstr /BR /C:"Rule Name" /C:"-" /C:"Enabled" /C:"Profiles" /C:"LocalPort" /C:"Action:.*Allow" /C:"^$"

# Example output:

Rule Name:                            Remote Event Log Management (NP-In)
----------------------------------------------------------------------
Enabled:                              No
Profiles:                             Domain
LocalPort:                            445
Action:                               Allow

This will do the equivalent of -B / -A in grep to search for rules based on LocalPort, but PowerShell is required:

netsh advfirewall firewall show rule name=all dir=in | Select-String "LocalPort:.*443" -Context 10,10
netsh advfirewall firewall show rule name=all dir=in | Select-String "Rule Name:.*SSH" -Context 10,10

Add inbound rules for RDP and SSH:

netsh advfirewall firewall add rule name="Remote Desktop" dir=in protocol=TCP localport=3389 action=allow
netsh advfirewall firewall add rule name="SSH" dir=in protocol=TCP localport=22 action=allow

Add outbound egress rules:

netsh advfirewall firewall add rule name="Egress 10.0.0.0/8" dir=out protocol=any action=block remoteip=10.0.0.0/8
netsh advfirewall firewall add rule name="Default Gateway" dir=out protocol=any remoteip=<gateway-ip> interfacetype=lan action=allow
netsh advfirewall firewall add rule name="Egress fc00::/7" dir=out protocol=any action=block remoteip=fc00::/7

Delete a rule called "Remote Desktop"

netsh advfirewall firewall delete rule name="Remote Desktop"

Delete all rules for local port tcp/80:

netsh advfirewall firewall delete rule name=all protocol=tcp localport=80

Reset Firewall (To Defaults)

netsh advfirewall reset

Netsh Port Proxy

https://www.sans.org/blog/pen-test-poster-white-board-cmd-exe-c-netsh-interface/

Windows Net Shell command

  • Using 0.0.0.0 as the listen address makes the portproxy bind to all interfaces
  • Using v4tov6 allows bidirectional IPv4 and IPv6 usage
netsh interface portproxy show all
netsh interface portproxy show v4tov4
netsh interface portproxy show v4tov6
netsh interface portproxy show v6tov4
netsh interface portproxy show v6tov6

netsh interface portproxy add v4tov4 listenaddress=<local-addr> listenport=<local-port> connectaddress=<dest-addr> connectport=<dest-port>
netsh interface portproxy add v4tov4 listenaddress=<jump-box-addr> listenport=<jump-box-port> connectaddress=<attacker-addr> connectport=<attacker-web-server-port>
netsh interface portproxy add v4tov4 listenport=8443 listenaddress=10.0.0.15 connectport=443 connectaddress=172.16.1.28
netsh interface portproxy add v4tov4 listenport=443 listenaddress=0.0.0.0 connectport=443 connectaddress=172.16.1.28

Forwards traffic hitting victim's / jump box's listenaddress:listenport to attacker's connectaddress:connectport

blockinboundalways / -AllowInboundRules False

To drop all inbound connections even if Windows has a default allow rule for the service:

#cmd.exe
netsh advfirewall set allprofiles firewallpolicy blockinboundalways,allowoutbound

#powershell
Set-NetFirewallProfile -AllowInboundRules False
  • On domain joined workstations, this will not disrupt connections to server file shares and stops lateral movement between workstations
  • Workstation typically should not need to talk to each other, with the server being the central point of authentication
  • On personal, non domain joined workstations this should be the default setting, and an absolute must for travel / using untrusted LAN and WiFi networks

Keep in mind on cloud instances of Windows in Azure / AWS / GCP this will likely lock you out of the machine

To enable this setting from within the GUI:

  • Network & Internet Settings > Status > Windows Firewall
  • For all three, Domain, Public, Private:
  • Set to On
  • Check Blocks all incoming connections, including those on the list of allowed apps.

To turn off this setting via the GUI:

  • Uncheck Blocks all incoming connections, including those on the list of allowed apps.

LLMNR

https://www.blackhillsinfosec.com/how-to-disable-llmnr-why-you-want-to/

Disable via GPO:

  • Group Policy Management
  • Forest: <DOMAIN>.local > Domains > <DOMAIN>.local > Right-Click
  • Create a GPO in this domain, and Link it here...
  • Name it "LLMNR" or something meaningful, click OK
  • Computer Configuration > Administrative Templates > Network > DNS Client
  • Turn Off Multicast Name Resolution > Enabled
  • Right-Click the policy name under <DOMAIN>.local and choose Enforced

You'll need to restart the endpoints, or refresh the group policy on each endpoint manually to ensure these changes take effect immediately

Update Group Policy via PowerShell:

Get-ADComputer -filter * | foreach{ Invoke-GPUpdate -computer $_.name -RandomDelayInMinutes 0 -force}

Update Group Policy via cmd.exe:

gpupdate

Disable LLMNR via PowerShell:

If (!(Test-Path "HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\DNSClient")) {
	New-Item -Path "HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\DNSClient" -Force | Out-Null
}
Set-ItemProperty -Path "HKLM:\SOFTWARE\Policies\Microsoft\Windows NT\DNSClient" -Name "EnableMulticast" -Type DWord -Value 0

Disable LLMNR via cmd.exe:

REG ADD  “HKLM\Software\policies\Microsoft\Windows NT\DNSClient”
REG ADD  “HKLM\Software\policies\Microsoft\Windows NT\DNSClient” /v ” EnableMulticast” /t REG_DWORD /d “0” /f

Mitigate IPv6 MitM Attacks

https://academy.tcm-sec.com/p/practical-ethical-hacking-the-complete-course

https://blog.fox-it.com/2018/01/11/mitm6-compromising-ipv4-networks-via-ipv6/

https://dirkjanm.io/worst-of-both-worlds-ntlm-relaying-and-kerberos-delegation/

There are three predefined firewall rules to change from allow to block.

This walks through setting these rules as a GPO on a Domain Controller, but you can create the same rules manually on endpoints via Windows Defender Firewall.

Test these rules in an environment with ADCS configured by enabling them, updating Group Policy on each machine, then rebooting endpoints and signing in as a Domain Admin on any of them with mitm6 running. The attack(s) will no longer fire when these rules are in place.

Here are the commands to run on the attacker machine:

# In one terminal window:
sudo /home/kali/.local/bin/mitm6 -i eth0 -v -d TESTDOMAIN.local
# In another:
impacket-ntlmrelayx -6 -t ldaps://<dc-ip> -wh fakewpad.testdomain.local -l loot
  • Group Policy Management

  • Forest: <DOMAIN>.local > Domains > <DOMAIN>.local > Right-Click

  • Create a GPO in this domain, and Link it here...

  • Name it "IPv6" or something meaningful, click OK Computer Configuration > Windows Settings > Security Settings > Windows Defender Firewall with Advanced Security > Windows Defender Firewall with Advanced Security > [Inbound|Outbound] Rules

  • Create new custom rules for all three, there are usually none defined in Group Policy Management on the DC when defining a new GPO.

  • (Inbound) Core Networking - Dynamic Host Configuration Protocol for IPv6(DHCPV6-IN)

    • Name: Block - (DHCPV6-In)
    • Protocol type: UDP
    • Protocol number: 17
    • Local Port: Specific Ports, 546
    • Remote Port: Specific Ports, 547
    • This program: %SystemRoot%\system32\svchost.exe
  • (Inbound) Core Networking - Router Advertisement (ICMPv6-IN)

    • Name: Block - Router Advertisement (ICMPv6-In)
    • Protocol type: ICMPv6
    • Protocol number: 58
    • Local port: All Ports
    • Remote port: All Ports
    • Specific ICMP types: [x] Router Advertisement
      • ICMP Type: 0
      • ICMP Code: Any
    • This program: System
  • (Outbound) Core Networking - Dynamic Host Configuration Protocol for IPv6 (DHCPV6-Out)

    • Name: Block - (DHCPV6-Out)
    • Protocol type: UDP
    • Protocol number: 17
    • Local Port: Specific Ports, 546
    • Remote Port: Specific Ports, 547
    • This program: %SystemRoot%\system32\svchost.exe
  • Right-Click the policy name under <DOMAIN>.local and choose Enforced

You'll need to restart the endpoints or refresh the group policy on each endpoint manually to ensure these changes take effect immediately

Update Group Policy via PowerShell:

Get-ADComputer -filter * | foreach{ Invoke-GPUpdate -computer $_.name -RandomDelayInMinutes 0 -force}

Update Group Policy via cmd.exe:

gpupdate

Enable SMB Signing

https://learn.microsoft.com/en-us/troubleshoot/windows-server/networking/overview-server-message-block-signing

Default for Workstations = 0 Default for Server = 1

Enable SMB Signing via PowerShell:

Set-ItemProperty -Path "HKLM:\System\CurrentControlSet\Services\LanManWorkstation\Parameters" -Name "RequireSecuritySignature" -Type DWord -Value 1
Set-ItemProperty -Path "HKLM:\System\CurrentControlSet\Services\LanManServer\Parameters" -Name "RequireSecuritySignature" -Type DWord -Value 1

Enable SMB Signing on a Domain Controller via a GPO:

  • Group Policy Management
  • Forest: <DOMAIN>.local > Domains > <DOMAIN>.local > Right-Click
  • Create a GPO in this domain, and Link it here...
  • Name it "SMB Signing" or something meaningful, click OK
  • Computer Configuration > Policies > Windows Settings > Security Settings > Local Policies > Security Options
  • Microsoft network server: Digitally sign communications (always) > Enable
  • Microsoft network client: Digitally sign communications (always) > Enable
  • Right-Click the policy name under <DOMAIN>.local and choose Enforced

You'll need to restart the endpoints or refresh the group policy on each endpoint manually to ensure these changes take effect immediately

Update Group Policy via PowerShell:

Get-ADComputer -filter * | foreach{ Invoke-GPUpdate -computer $_.name -RandomDelayInMinutes 0 -force}

Update Group Policy via cmd.exe:

gpupdate

Filesystem Permissions

How to do the equivalent of chmod operations in Windows.

takeown.exe

Change ownership recursively of a file(s) or folder(s) to the current user:

takeown.exe /F .\ExampleDir\ /R

Change ownership recursively of a file(s) or folder(s) to the Administrators group:

takeown.exe /F .\ExampleDir\ /A /R

Change ownership back to a standard user (where $HOSTNAME is either local pc name or domain name, and user is a local or domain user):

takeown.exe /S $HOSTNAME /U $USER /F .\ExampleDir\

icacls.exe

Remove all permissions on a file or folder:

icacls.exe .\ExampleDir /inheritance:r

Reset to default permissions:

icacls.exe .\ExampleDir\ /reset

Folder Permissions

Grant read-only access to a folder for user alice:

icacls.exe .\ExampleDir\ /grant alice:"(CI)(OI)RX"

Remove granted (note the :g) access to a *folder for user alice:

icacls.exe .\ExampleDir\ /remove:g alice

Permit read-only access to a folder for the builtin group "everyone" (removes any inherit or current write or modify permissions with /inheritance:r):

  • SID *S-1-1-0 means everyone
  • Similar to chmod a=rX -R ./ExampleFolder in Linux
icacls.exe .\ExampleDir\ /inheritance:r
icacls.exe .\ExampleDir\ /grant *S-1-1-0:"(CI)(OI)RX"

The purpose of CI and OI is to allow the "synchronize" permission, which allows directory traversal and if missing, denies it even if RX permission is granted

Configure a folder to be only modifiable by administrators, read and execute by eveyrone:

  • Similar to chmod a=rX -R ./ExampleFolder; chmod o=rwX -R ./ExampleFolder; chown root:root ./ExmapleFolder on Linux
icacls.exe C:\Tools /inheritance:r
icacls.exe C:\Tools /grant *S-1-1-0:"(CI)(OI)RX"
icacls.exe C:\Tools /grant SYSTEM:"(CI)(OI)(F)"
icacls.exe C:\Tools /grant BUILTIN\Administrators:"(CI)(OI)(F)"

Example output:

PS> icacls.exe C:\Tools\*
C:\Tools Everyone:(OI)(CI)(RX)
         NT AUTHORITY\SYSTEM:(OI)(CI)(F)
         BUILTIN\Administrators:(OI)(CI)(F)

File Permissions

Grant read-only access to a file for user alice. This does not require (CI)(OI):

icacls.exe .\ExampleFile /grant alice:"RX"

Remove granted (note the :g) access to a file for user alice:

icacls.exe .\ExampleFile /remove:g alice

Permit read-only + execute for the everyone builtin group, to a single file. This does not require (CI)(OI):

icacle.exe .\example.md /inheritance:r /grant *S-1-1-0:"RX"

Limit full access to only SYSTEM and BUILTIN\Administrators:

icacls.exe .\administrators_authorized_keys /inheritance:r
icacls.exe .\administrators_authorized_keys /grant SYSTEM:"(F)"
icacls.exe .\administrators_authorized_keys /grant BUILTIN\Administrators:"(F)"

Example output:

PS> icacls.exe .\administrators_authorized_keys
administrators_authorized_keys NT AUTHORITY\SYSTEM:(F)
                               BUILTIN\Administrators:(F)

Set-Acl

Take the ACL data of one filesystem object and apply it to another

$NewAcl = Get-Acl -Path C:\Users\Administrator\Documents
Set-Acl -Path C:\Tools -AclObject $NewAcl

To use this like icacls.exe we'll need to reference these examples:

Looking at the three arguments to FileSystemAccessRule($identity, $fileSystemRights, $type), shows how we can start to write the code block.

Aside from the list of fields for $fileSystemRights, $identity is a reference to a user account and $type is either Allow (0) or Deny (1).

SMB

This section references the following sources:

Create a new folder, an SMB share for that folder, and give flare-user full access to it:

# Setup a user to access the folder
$Password = ConvertTo-SecureString "flare-pass" -AsPlainText -Force
New-LocalUser "flare-user" -Password $Password
Add-LocalGroupMember -Group "Users" -Member "flare-user"

# Create the folder
New-Item -Type Directory -Path C:\Share

# Give flare-user FullControl
$NewAcl = Get-Acl -Path "C:\Share"
$identity = "[DOMAIN\]flare-user"
$fileSystemRights = "FullControl"
$type = "Allow"
$argumentlist = $identity, $fileSystemRights, $type
$NewRule = New-Object -TypeName System.Security.AccessControl.FileSystemAccessRule -ArgumentList $argumentlist
$NewAcl.SetAccessRule($NewRule)
Set-Acl -Path "C:\Share" -AclObject $NewAcl

# Create the share
New-SmbShare -Path C:\Share -Name "Share" -FullAccess "flare-user"

# Check available shares
Get-SmbShare -Name *

NOTE 1: You can tab complete difficult to remember arguments like System.Security.AccessControl.FileSystemAccessRule.

NOTE 2: ChatGPT May 24 Version demonstrates a shorter way of writing the $NewRule line:

$NewRule = New-Object -TypeName System.Security.AccessControl.FileSystemAccessRule("flare-user", "FullControl", "Allow")

From here you can easily impacket-smbclient flare-user:flare-pass@hostname to connect.

Modifying access to a share is done with the following cmdlets:

Remove an SMB share:

Remove-SmbShare -Name "Share"

SysInternals

Overview:

https://docs.microsoft.com/en-us/sysinternals/

Download individual tools directly:

https://live.sysinternals.com/

Download the entire suite as a zip archive:

https://download.sysinternals.com/files/SysinternalsSuite.zip

AccessChk

See the following blog for additional uses of AccessChk:

Evaluate what access to C:$PATH is available to $USER:

accesschk64.exe "$USER" c:\$PATH

Sysmon

Installing Sysmon

Open an administrative PowerShell session.

Create the tools directory with the correct permissions:

New-Item -ItemType Directory -Path "C:\Tools" | Out-Null
New-Item -ItemType Directory -Path "C:\Tools\Scripts" | Out-Null
icacls.exe C:\Tools /reset | Out-Null
icacls.exe C:\Tools /inheritance:r | Out-Null
icacls.exe C:\Tools /grant SYSTEM:"(CI)(OI)(F)" | Out-Null
icacls.exe C:\Tools /grant BUILTIN\Administrators:"(CI)(OI)(F)" | Out-Null
icacls.exe C:\Tools /grant *S-1-1-0:"(CI)(OI)RX" | Out-Null

Download and install Sysmon and a starter configuration file (if you did not write your own).

# Change to the tools path
cd C:\Tools

# If you want to use SwiftOnSecurity's config:
iwr "https://raw.githubusercontent.com/SwiftOnSecurity/sysmon-config/master/sysmonconfig-export.xml" -OutFile "C:\Tools\sysmon-config.xml"
# If you want to use Olaf Hartong's config:
iwr "https://raw.githubusercontent.com/olafhartong/sysmon-modular/master/sysmonconfig.xml" -OutFile "C:\Tools\sysmon-config.xml"

# Download Sysmon
iwr "https://live.sysinternals.com/Sysmon64.exe" -OutFile "C:\Tools\Sysmon64.exe"

# Uninstall the currently running Sysmon components with the newest binary if you already have an older version running (does not require reboot)
C:\Tools\Sysmon64.exe -accepteula -u

# Install the newest Sysmon (does not require reboot)
C:\Tools\Sysmon64.exe -accepteula -i C:\Tools\sysmon-config.xml

NOTE: This does not erase or remove current log files, and they can all still be read again after installing the new binary.

Cleanup

  • Option 1: Make the config file readable only by SYSTEM and BUILTIN\Administrator
     icacls.exe C:\Tools\sysmon-config.xml /inheritance:r
     icacls.exe C:\Tools\sysmon-config.xml /grant SYSTEM:"(F)"
     icacls.exe C:\Tools\sysmon-config.xml /grant BUILTIN\Administrators:"(F)"
  • Option 2: Delete the config file from the local machine
  • Both: Monitor and log for execution of Sysmon64.exe -c which dumps the entire configuration whether it's still on disk or not. If you find this in your logs and did not run this, you may have been broken into.

Custom Config

Using Sysmon-Modular it's easy to create custom configuration files.

Install the configuration with C:\Tools\Sysmon64.exe -c .\sysmonconfig.xml and observe it with your SIEM, the Event Viewer, or Tail-EventLogs.ps1.

Adjust rules accordingly. For example, if there's a DLL Side-Loading rule that's overwhelming your logs, you can find which rule files include this rule using the following:

. .\Merge-AllSysmonXml

Find-RulesInBasePath -BasePath C:\Tools\sysmon-modular\ | sls "DLL Side-Loading"
  • Read the rule criteria based on what you're seeing in your logs.
  • Find the exact line in sysmon-modular's rules to tune it manually.
  • Regenerate the config using Merge-AllSysmonXml -Path ( Get-ChildItem '[0-9]*\*.xml') -AsString | Out-File sysmonconfig.xml.

Tune a Config (with PowerShell)

This will depend on your configuration file. The sysmon-modular config mentioned above is a great starting place to help you maintain includes and excludes.

When first applying a configuration, do it in a test environment mirroring the systems it will be deployed on. Then review what's being logged that's either too much, or not enough. This will likely be accomplished with a SIEM, but if you need to do this with PowerShell you can investigate logs by working big to small in scope.

  • This will help identify what's being logged too much
  • This is not the most effective way to find what isn't being logged
  • To find what's not being logged you'll need to conduct (offensive) testing on your environment

Find the frequency of all Sysmon Events:

$StartDate = (Get-Date).AddDays(-1)
Get-WinEvent -FilterHashtable @{ Logname='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate } | Group-Object -Property Id | Select-Object Name, Count | Sort-Object Count -Descending

Find the frequency of a known property in each event (in this case, the RuleName from sysmon-modular):

$StartDate = (Get-Date).AddDays(-1)
Get-WinEvent -FilterHashtable @{ Logname='Microsoft-Windows-Sysmon/Operational'; Id='10' } | ForEach-Object { $_.properties[0].value }| Group-Object | Sort-Object -Property Count, Descending | select Count, Name

Windows Logging

To see all available logs on a (Windows) system:

Get-WinEvent -ListLog * | Select -Property LogName

To see the available properties for a log, use:

Get-WinEvent -MaxEvents 1 -LogName '<log>' | select -Property *

Event Logs

Log Size

Independant of how long your SIEM or central logging server is retaining all logs shipped to it, each endpoint should maintain between 200MB to 1GB (or more if the machine can tolerate it) of Sysmon and or Security event logs to ensure all events are properly captured and forwarded (not lost).

Windows Defender Logs

Get the latest instance of Id 1116, MALWAREPROTECTION_STATE_MALWARE_DETECTED:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Windows Defender/Operational'; StartTime=$StartDate; Id='1116'; } | Select -First 1 | fl

Get all "Warning" events that aren't Id 1002, MALWAREPROTECTION_SCAN_CANCELLED (which can be a noisy event):

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Windows Defender/Operational'; StartTime=$StartDate; } | where { $_.LevelDisplayName -eq 'Warning' -and $_.Id -ne 1002 }

Find any instances of Defender's configuration being changed:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Windows Defender/Operational'; StartTime=$StartDate; } | where { $_.Id -eq 5004 -or  $_.Id -eq 5007  }

Find any instances of Defender components being disabled or off:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Windows Defender/Operational'; StartTime=$StartDate; } | where { $_.Id -eq 5001 -or  $_.Id -eq 5008 -or $_.Id -eq 5010 -or $_.Id -eq 5012 }

ASR Events

The link above includes raw XML you can import into the event viewer to filter for only ASR related log entries. The XML also contains the paths needed to query the event logs from PowerShell. This is a copy of that table for reference:

Feature Provider/source Event ID Description
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 1 ACG audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 2 ACG enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 3 Don't allow child processes audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 4 Don't allow child processes block
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 5 Block low integrity images audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 6 Block low integrity images block
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 7 Block remote images audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 8 Block remote images block
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 9 Disable win32k system calls audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 10 Disable win32k system calls block
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 11 Code integrity guard audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 12 Code integrity guard block
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 13 EAF audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 14 EAF enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 15 EAF+ audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 16 EAF+ enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 17 IAF audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 18 IAF enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 19 ROP StackPivot audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 20 ROP StackPivot enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 21 ROP CallerCheck audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 22 ROP CallerCheck enforce
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 23 ROP SimExec audit
Exploit protection Security-Mitigations (Kernel Mode/User Mode) 24 ROP SimExec enforce
Exploit protection WER-Diagnostics 5 CFG Block
Exploit protection Win32K (Operational) 260 Untrusted Font
Network protection Windows Defender (Operational) 5007 Event when settings are changed
Network protection Windows Defender (Operational) 1125 Event when Network protection fires in Audit-mode
Network protection Windows Defender (Operational) 1126 Event when Network protection fires in Block-mode
Controlled folder access Windows Defender (Operational) 5007 Event when settings are changed
Controlled folder access Windows Defender (Operational) 1124 Audited Controlled folder access event
Controlled folder access Windows Defender (Operational) 1123 Blocked Controlled folder access event
Controlled folder access Windows Defender (Operational) 1127 Blocked Controlled folder access sector write block event
Controlled folder access Windows Defender (Operational) 1128 Audited Controlled folder access sector write block event
Attack surface reduction Windows Defender (Operational) 5007 Event when settings are changed
Attack surface reduction Windows Defender (Operational) 1122 Event when rule fires in Audit-mode
Attack surface reduction Windows Defender (Operational) 1121 Event when rule fires in Block-mode

Notable paths for PowerShell parsing include the following (keeping in mind only the event ID's above apply):

  • LogName='Microsoft-Windows-Security-Mitigations/UserMode'
  • LogName='Microsoft-Windows-Security-Mitigations/KernelMode'
  • LogName='Microsoft-Windows-Win32k/Operational'
  • LogName='Microsoft-Windows-WER-Diag/Operational'
  • LogName='Microsoft-Windows-Windows Defender/Operational'

Example queries, based on LogName and $_.Id:

$StartDate = (Get-Date).AddDays(-1)
Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Security-Mitigations/UserMode'; StartTime=$StartDate; }
Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Security-Mitigations/KernelMode'; StartTime=$StartDate; } | where { $_.Id -neq 11 -and  $_.Id -neq 12  }
Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Windows Defender/Operational'; StartTime=$StartDate; } | where { $_.Id -eq 1121 -or  $_.Id -eq 1122  } | fl *

It's important to note that when a user is notified via a toast notificaiton, the alert details are not always available within the Windows Defender GUI to review. Using PowerShell will allow you to extract any matching logs and all details.

Logon Events

This can be difficult to understand at first, as the normal Logon (ID 4624) and Logoff (ID 4634) events will produce numerous entries even if you simply sign out and back in to an account locally. Using these events alone can be difficult if you're trying to trace behavior. If you're looking for something similar to Linux's last | head then you need to query the Explicit Logon Attempt event (ID 4648).

Get-WinEvent -FilterHashtable @{LogName='Security'} | where TimeCreated -gt (Get-Date).AddDays(-1) | where Id -eq '4648'
Get-WinEvent -FilterHashtable @{LogName='Security'} | where TimeCreated -gt (Get-Date).AddDays(-1) | where Id -eq '4648' | fl

You can also use offensive tools to do this, such as Seatbelt, which returns easily readable results:

.\Seatbelt.exe ExplicitLogonEvents

I wanted to mimic Seathbelt's output in PowerShell, so I wrote this script to accomplish that. This includes references to the Sysmon Logs section here, and also to the C# source code for the ExplicitLogonEvents module in Seatbelt.

# MIT License
#
# Print unique logon information for a Windows system using built in event logs, in a clear way.
# This returns every user login by default, but you can swap around variables below to suit your needs
# As is, this will also catch shells like "runas /user:username cmd.exe"
#
# This script is meant to return data in a similar format as ".\Seatbelt.exe ExplicitLogonEvents" would:
# [GhostPack/Seatbelt - ExplicitLogonEvents](https://github.com/GhostPack/Seatbelt/blob/master/Seatbelt/Commands/Windows/EventLogs/ExplicitLogonEvents/ExplicitLogonEventsCommand.cs)
#
# Parsing logs like this was learned from:
# [Security Weekly Webcast - Sysmon: Gaining Visibility Into Your Enterprise](https://securityweekly.com/webcasts/sysmon-gaining-visibility-into-your-enterprise/)
# [Slide Deck](https://securityweekly.com/wp-content/uploads/2022/04/alan-stacilauskas-unlock-sysmon-slide-demov4.pdf)
#
# Huge thanks to the presentors:
#
# [Alan Stacilauskas](https://www.alstacilauskas.com/)
# [Amanda Berlin](https://twitter.com/@infosystir)
# [Tyler Robinson](https://twitter.com/tyler_robinson)

# Set a range of time to search in the logs
# AddMinutes / AddHours / AddDays / AddMonths
$startTime = (Get-Date).AddDays(-1)

# This line will get you the relevant event log data to begin parsing
Get-WinEvent -FilterHashtable @{LogName='Security'} | where TimeCreated -gt $startTime | where Id -eq '4648' | foreach {
    # Variable names were kept the same as they are in Seatbelt's ExplicitLogonEventsCommand.cs
    # Variable properties can be extracted from the Windows Event ID 4648 $_.Message object:
    $creationTime = $_.TimeCreated
    $subjectUserSid = $_.properties[0].value
    $subjectUserName = $_.properties[1].value
    $subjectDomainName = $_.properties[2].value
    $subjectLogonId = $_.properties[3].value
    $logonGuid = $_.properties[4].value
    $targetUserName = $_.properties[5].value
    $targetDomainName = $_.properties[6].value
    $targetLogonGuid = $_.properties[7].value
    $targetServerName = $_.properties[8].value
    $targetServerInfo = $_.properties[9].value
    $processId = $_.properties[10].value
    $processName = $_.properties[11].value
    $ipAddress = $_.properties[12].value
    $ipPort = $_.properties[13].value

    # You can change this pattern variable and the if statement itself below if you'd like to search for different data
    # Basically the "user logging in" part of the login sequence captured by event logs is made by svchost.exe and not winlogon.exe
    $patternToMatch = 'winlogon.exe'

    # Another approach is filtering for a blank username:
    # if ($targetUserName -ne $nul) { ...
    #
    # Alternatively you could filter for:
    # if ($targetUserName -inotmatch "(UMFD-\d|DWM-\d)") { ...
    # DWM-\d matches patterns of the Desktop Window Manager user
    # UMFD-\d matches patterns of the Font Driver Host user

    # Lastly, you could filter based on the source IP Address:
    # if ($ipAddress -match $patternToMatch) {

    if ($processName -inotmatch $patternToMatch) {
        Write-Host "$creationTime, $targetUserName, $targetDomainName, $targetServerName, $processName, $ipAddress"
    }
}

PowerShell Logs

  • At minimum, enable Script Block logging (without Invocation Logging)
  • If possible, enable Transcription to a protected directory that's sent to a central logging server

Module Logging

Enable logging for selected PowerShell modules.

This may not be necessary if you enable script block logging.

Script Block Logging

Logs PowerShell input (commands, functions, decodes and tracks script content)

  • Enable Script Block Logging
  • Get-ItemProperty -Path "HKLM:\Software\Policies\Microsoft\Windows\PowerShell\ScriptBlockLogging"
  • Computer Configuration > Administrative Templates > Windows Components > Windows PowerShell > Turn on PowerShell Script Block Logging
  • Log Name: Microsoft-Windows-PowerShell/Operational
  • Log Path: C:\Windows\System32\winevt\Logs\Microsoft-Windows-PowerShell%4Operational.evtx

Invocation Logging is an entirely optional checkbox under this Group Policy setting. It logs the start and stop of executions. This will flood your logs.

Enable Script Block Logging (without the start / stop Invocation Logging option):

# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_logging?view=powershell-5.1#using-the-registry
$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell\ScriptBlockLogging'
) -join '\'

if (-not (Test-Path $basePath)) {
    $null = New-Item $basePath -Force
}

Set-ItemProperty $basePath -Name EnableScriptBlockLogging -Value "1"
Set-ItemProperty $basePath -Name EnableScriptBlockInvocationLogging -Value "0"

Remove Script Block Logging:

# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_logging?view=powershell-5.1#using-the-registry
$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell\ScriptBlockLogging'
) -join '\'

if (Test-Path $basePath) {
    Remove-ItemProperty $basePath -Name EnableScriptBlockLogging
    Remove-ItemProperty $basePath -Name EnableScriptBlockInvocationLogging
}

Script Block logging has the benefit of being able to unwrap obfuscated functions and record their raw text.

This code sample from the Microsoft Docs link above reassembles large script blocks logged across multiple entries:

$created = Get-WinEvent -FilterHashtable @{ ProviderName="Microsoft-Windows-PowerShell"; Id = 4104 } |
  Where-Object { $_.<...> }
$sortedScripts = $created | sort { $_.Properties[0].Value }
$mergedScript = -join ($sortedScripts | % { $_.Properties[2].Value })

Transcription Logging

Logs PowerShell output (everything that appears in the powershell terminal session)

  • PowerShell Transcription
  • HKLM:\Software\Policies\Microsoft\Windows\PowerShell\Transcription
  • Computer Configuration > Administrative Templates > Windows Components > Windows PowerShell > Turn on PowerShell Script Block Logging
  • Log path is up to the user, written as a text file

Enable PowerShell transcription, write output to C:\PSTranscripts:

# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_logging?view=powershell-5.1#using-the-registry
# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_powershell_config?view=powershell-7.3#transcription
# https://github.com/clr2of8/PowerShellForInfoSec/blob/main/Tools/Set-PSLogging.ps1
$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell\Transcription'
) -join '\'

if (-not (Test-Path $basePath)) {
    $null = New-Item $basePath -Force
}

Set-ItemProperty $basePath -Name EnableTranscripting -Value "1"
Set-ItemProperty $basePath -Name EnableInvocationHeader -Value "1"
Set-ItemProperty $basePath -Name OutputDirectory -Value "C:\PSTranscripts"

Remove PowerShell Transcription:

# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_logging?view=powershell-5.1#using-the-registry
# https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_powershell_config?view=powershell-7.3#transcription
# https://github.com/clr2of8/PowerShellForInfoSec/blob/main/Tools/Set-PSLogging.ps1
$basePath = @(
    'HKLM:\Software\Policies\Microsoft\Windows'
    'PowerShell\Transcription'
) -join '\'

if (Test-Path $basePath) {
    Remove-ItemProperty $basePath -Name EnableTranscripting
    Remove-ItemProperty $basePath -Name EnableInvocationHeader
    Remove-ItemProperty $basePath -Name OutputDirectory
}

Sysmon Logs

This is a quick start on how to read your Sysmon logs.

These will help in building statements to parse logs conditionally with more granularity:

Note that if you do not use Sort-Object -Unique or similar, logs will be displayed from oldest (top) to newest (bottom). It is also not necessary to use Out-String if you prefer working with PowerShell objects over strings.

Another tip is using <cmds>... | Group-Object | Select-Object Count, Name | Sort-Object Count -Descending. This is the PowerShell equivalent to Unix's ... | sort | uniq -c | sort -nr, grouping unique results into order based on occurrance.

This webcast is an excellent resource, not just for rule creation but various ways to script and parse logs, and essentially a quick start to Sysmon threat hunting via the CLI and GUI:

Huge thanks to the presentors:

Additional resources from the presentation, discord, and other trainings:

There are two easy ways of parsing log output:

Technique TypeName
| select { $_.properties[4].value } Selected.System.Diagnostics.Eventing.Reader.EventLogRecord
| ForEach-Object { Out-String -InputObject $_.properties[4].value } System.String

The technique of using ForEach-Object { Out-String -InputObject $_.properties[x,y,z].value } was highlighted during the webcast.


To parse any event logs quickly and effectively:

You can get any log's property values by piping it to | select -First 1 | fl. Each line in the Message block relates to a value.

Specify multiple values like this: $_.properties[0,1,5,6].value

You can also specifiy a series of values with two .. dots between them like: $_.properties[0..20].value

Set the start time as a variable:

$StartDate = (Get-Date).AddMinutes(-30)
$StartDate = (Get-Date).AddHours(-1)
$StartDate = (Get-Date).AddDays(-2)

Sysmon Event IDs

These references can help you sort through event properties visually.

ID 22: DNS Queries

DNS query property values:

  • [0]RuleName
  • [1]UtcTime
  • [2]ProcessGuid
  • [3]ProcessId
  • [4]QueryName
  • [5]QueryStatus
  • [6]QueryResults
  • [7]Image
  • [8]User

Show all unique DNS queries (ID 22) and sort them by frequency:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='22' } | ForEach-Object { $_.properties[4].value } | Group-Object | Sort-Object -Property Count -Descending | Select Count,Name

Show all DNS queries (ID 22) that contain "google":

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='22' } | Where-Object { $_.properties[4].value -imatch "google" } | ForEach-Object { $_.properties[4].value } | Group-Object | Sort-Object -Property Count -Descending | Select Count,Name

Show all DNS queries (ID 22) and when they were made:

Get-WinEvent -FilterHashtable @{ Logname='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='22' } | ForEach-Object { Out-String -InputObject $_.properties[1,4].value }

ID 3: Network Connection

Network connection property values:

  • [0]RuleName
  • [1]UtcTime
  • [2]ProcessGuid
  • [3]ProcessId
  • [4]Image
  • [5]User
  • [6]Protocol
  • [7]Initiated
  • [8]SourceIsIpv6
  • [9]SourceIp
  • [10]SourceHostname
  • [11]SourcePort
  • [12]SourcePortName
  • [13]DestinationIsIpv6
  • [14]DestinationIp
  • [15]DestinationHostname
  • [16]DestinationPort
  • [17]DestinationPortName

Show all unique executables that made a network connection, filter out msedge.exe and svchost.exe, sort by frequency:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; Id='3' } | where { $_.properties[4].value -inotmatch "(msedge.exe|svchost.exe)" } | ForEach-Object { $_.properties[4].value } | Group-Object | Sort-Object -Property Count -Descending | Select Count, Name

Using the query above, get all unique destination IPs, sort by frequency:

  • Use this to filter network activity
  • Feed the returned list of IPs to a threat intel platform
Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; Id='3' } | where { $_.properties[4].value -inotmatch "(msedge.exe|svchost.exe)" } | ForEach-Object { $_.properties[14].value } | Group-Object | Sort-Object -Property Count -Descending | Select Count, Name

Show all network connections (ID 3), what executable made them, when, and destination IP / hostname:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='3' } | ForEach-Object { Out-String -InputObject $_.properties[1,4,14,15].value }

ID 1: Process Creation

Process creation property values:

  • [0]RuleName
  • [1]UtcTime
  • [2]ProcessGuid
  • [3]ProcessId
  • [4]Image
  • [5]FileVersion
  • [6]Description
  • [7]Product
  • [8]Company
  • [9]OriginalFileName
  • [10]CommandLine
  • [11]CurrentDirectory
  • [12]User
  • [13]LogonGuid
  • [14]LogonId
  • [15]TerminalSessionId
  • [16]IntegrityLevel
  • [17]Hashes
  • [18]ParentProcessGuid
  • [19]ParentProcessId
  • [20]ParentImage
  • [21]ParentCommandLine
  • [22]ParentUser

List all processes created (ID 1) by timestamp, PID, executable, commandline, executable hashes, and PPID:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='1' } | ForEach-Object { Out-String -InputObject $_.properties[1,3,4,10,17,19].value }

List all details of all processes created (ID 1):

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='1' } | fl

List all details of (ID 1) log entries where the RuleName field contains "Discovery":

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='1' } | where { $_.properties[0].value -imatch "Discovery" } | fl

List how many times each executable created a process (ID 1), sorted by frequency:

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='1' } | ForEach-Object { $_.properties[4].value } | Group-Object | Sort-Object -Property Count -Descending | Select Count,Name

ID 10: Process Accessed

Process accessed property values:

  • [0]RuleName
  • [1]UtcTime
  • [2]SourceProcessGUID
  • [3]SourceProcessId
  • [4]SourceThreadId
  • [5]SourceImage
  • [6]TargetProcessGUID
  • [7]TargetProcessId
  • [8]TargetImage
  • [9]GrantedAccess
  • [10]CallTrace
  • [11]SourceUser
  • [12]TargetUser

Detect LSASS access (currently this just matches the first entry that has the string "lsass" anywhere in the log, needs revised to be more precise):

Get-WinEvent -FilterHashtable @{ LogName='Microsoft-Windows-Sysmon/Operational'; StartTime=$StartDate; Id='10' } | Where-Object { $_.properties[0..30].value -imatch "lsass" } | Select -First 1 | fl *

Services

OpenSSH

Directory Description
C:\ProgramData\ssh\ Main configuration folder, /etc/ssh/ equivalent
C:\Users\$USER\.ssh\ User folder, ~/.ssh equivalent

Steps to setup OpenSSH Server (The OpenSSH Client is typically installed by default).

  1. Update Windows

  2. Install OpenSSH Server Optional Feature

See the following Microsoft documentation of OpenSSH for additional context:

https://docs.microsoft.com/en-us/windows-server/administration/openssh/openssh_install_firstuse

Get-WindowsCapability -Online | Where-Object Name -like 'OpenSSH*'

# Install the OpenSSH Client
Add-WindowsCapability -Online -Name OpenSSH.Client~~~~0.0.1.0

# Install the OpenSSH Server
Add-WindowsCapability -Online -Name OpenSSH.Server~~~~0.0.1.0

# Start the sshd service
Start-Service sshd
  1. Configure:
# OPTIONAL but recommended:
Set-Service -Name sshd -StartupType 'Automatic'

# Confirm the Firewall rule is configured. It should be created automatically by setup. Run the following to verify
if (!(Get-NetFirewallRule -Name "OpenSSH-Server-In-TCP" -ErrorAction SilentlyContinue | Select-Object Name, Enabled)) {
    Write-Output "Firewall Rule 'OpenSSH-Server-In-TCP' does not exist, creating it..."
    New-NetFirewallRule -Name 'OpenSSH-Server-In-TCP' -DisplayName 'OpenSSH Server (sshd)' -Enabled True -Direction Inbound -Protocol TCP -Action Allow -LocalPort 22
} else {
    Write-Output "Firewall rule 'OpenSSH-Server-In-TCP' has been created and exists."
}

Change the following in C:\ProgramData\ssh\sshd_config:

#PasswordAuthentication Yes

To:

PasswordAuthentication no

Place public key data into:

C:\Users\$USER\.ssh\authorized_keys
C:\ProgramData\ssh\administrators_authorized_keys

Like on Unix systems, the permissions must be correct on the authorized_key files. See the following page on the Win32-OpenSSH Wiki for detailed descriptions:

https://github.com/PowerShell/Win32-OpenSSH/wiki/Security-protection-of-various-files-in-Win32-OpenSSH

Then, if the user you'll be logging in as IS NOT an administrator:

takeown.exe /F .\authorized_keys /S $HOSTNAME /U $USER
icacls.exe .\authorized_keys /reset
icacls.exe .\authorized_keys /inheritance:r
icacls.exe .\authorized_keys /grant $USER:"(F)"

or if the user you'll be logging in as IS an administrator:

icacls.exe .\authorized_keys /reset
icacls.exe .\administrators_authorized_keys /inheritance:r
icacls.exe .\administrators_authorized_keys /grant SYSTEM:"(F)"
icacls.exe .\administrators_authorized_keys /grant BUILTIN\Administrators:"(F)"
  1. Applying Configuration Changes

Anytime you update or modify C:\ProgramData\ssh\sshd_config, be sure to restart the sshd service so that it reads and loads the latest changes:

Restart-Service sshd

Scheduled Tasks

List all scheduled tasks by creation date:

Get-ScheduledTask -TaskName * | Select -Property Date,Author,Taskname | Sort-Object -Property Date

View scheduled tasks' creation date in the GUI:

Task Scheduler > Task Scheduler (Local) > Task Scheduler Library > [TaskName] > Created

Autoruns will also quickly identify any scheduled tasks in the Scheduled Tasks tab.

  • Hide: Windows Entries
  • Hide: Microsoft Entries

https://docs.microsoft.com/en-us/sysinternals/downloads/autoruns

Log & Audit Task Creation

https://docs.microsoft.com/en-us/windows/security/threat-protection/auditing/event-4698#security-monitoring-recommendations

Using the GUI:

Run gpedit.msc

Local Computer Policy > Windows Settings > Security Settings > Advanced Audit Policy Configuration > Object Access > Audit Other Object Access Events

Choose Configure the following audit events

Choose Success, then Apply and OK

Now you can query the event logs for occurances of scheduled task creation:

Get-WinEvent -LogName "Security" | Where Id -eq "4698"
Get-WinEvent -LogName "Security" | Where Id -eq "4698" | Select -Property *

TO DO: Configure scheduled task creation auditing using only PowerShell

You can locate registry items just like searching the filesystem.

We know the policy for scheduled task logging is related to "Object Access":

Get-ChildItem -Path "HKLM:\SOFTWARE\Microsoft\*object*access*" -Recurse -ErrorAction SilentlyContinue

After a possible delay while it searches, you should find it at this path, and be able to check it's values:

Get-ItemProperty -Path "HKLM:\SOFTWARE\Microsoft\PolicyManager\default\Audit\ObjectAccess_AuditOtherObjectAccessEvents"

WMI


Disk Management

This SuperUser answer by user VainMan is an excellent walkthrough of managing disks with just built in Windows tools:

PowerShell Cmdlets:

Additional documentation for this section:

Mount and Unmount Volumes and Drives (GUI)

The default external storage removal policy in recent versions of Windows (since Windows 10 1809) is Quick removal.

See: diskmgmt.msc > Right-Click the Disk number, select Properties > Policies Tab

  • Quick removal: You can remove the device without using the "Safely Remove Hardware" process
  • Better performance: Improved system resource usage, data is cached but you will need to use "Safely Remove Hardware" to avoid data loss

Set a Drive to Offline:

  • Also in the diskmgmt.msc window, Right-Click the Disk number
  • Select Offline to take the drive offline
  • This unmounts the drive, and prevents accidental writes or access
  • You can also use Set-Drive to take a disk offline or make it read-only

Mount and Unmount Volumes and Drives (CLI)

This is functionally the equivalent of using diskmgmt.mmc in any of the following ways:

  • Create a new simple volume from unallocated space, or format an external drive and assigning it a drive letter
  • Unmounting the drive (removing the drive letter)
  • Deleting the volume altogether

Get the VolumeName of the G: drive:

mountvol.exe G: /L
     \\?\Volume{12345678-abcd-efab-cdef-01234567890a}\

Unmount the G: drive

mountvol.exe G: /P
  • The volume will still exist
  • The mount point will no longer be traversable since there's no longer an assigned drive letter
  • Get-ItemProperty -Path "HKLM:\SYSTEM\MountedDevices" will show a #{GUID} instead of \DosDevices\G:
  • Even if you assign the volume a new drive letter, it will know it's tied to the same #{GUID} and replace it with \DosDevices\<letter>:

Mount the volume to a new drive letter (quote the VolumeName string):

mountvol.exe X: '\\?\Volume{12345678-abcd-efab-cdef-01234567890a}\'

Remove the volume, delete the volume (meaning all data on the volume), then clean up any leftover artifiacts in the registry

  • Get-ItemProperty -Path "HKLM:\SYSTEM\MountedDevices" will still show a #{GUID} after deleting a volume with any disk tool
  • The deleted volume should still be visible in diskmgmt.mmc until you delete its registry entry, then close and open diskmgmt.mmc again
  • mountvol.exe /R will remove this registry entry once the volume no longer exists
  • If the volume still exists, mountvol.exe /R will not remove the entry
  • If you format the same volume again and mount it, you'll note it retains the same #{GUID} when unmounted
mountvol.exe G: /P
diskpart.exe
> list volume
> select Volume 1
> delete
> exit
mountvol.exe /R

Extend the C Drive

Extending the C drive isn't intuitive because the partition the OS is installed on is sandwiched between the boot and recovery partitions. Even if you have free space (say on a newly extended VM or on a dual boot machine where you're removing the other OS) you cannot extend the installed partition without moving the recovery partition.

This entire example essentially mirrors the SuperUser answer linked above, with only small additions or changes to focus on backing up the recovery partition, deleting it so that the C: partition can be extended, then appending the backup recovery partition image to the new end of the C: partition.

Mount the recovery partition

diskpart
list disk
select disk <disk>
list partition
select partition <partition>
assign letter=R
exit

Create an image file with the dd-like dism.exe utility:

dism /Capture-Image /ImageFile:C\recovery-partition.wim /CaptureDir:R:\ /Name:"Recovery"

Delete the current recovery partition:

diskpart
select volume R
delete partition override
exit
  • Extend the C:\ drive in Disk Management
  • Create a new volume (+ new drive letter)
  • Apply the recovery image to the new volume using the new drive letter
dism /Apply-Image /ImageFile:C:\recovery-partition.wim /Index:1 /ApplyDir:R:\

Register the new recovery location

reagentc /disable
reagentc /setreimage /path R:\Recovery\WindowsRE
reagentc /enable

Unmount that drive letter to return the recovery partition to a 'hidden' state

diskpart
select volume R
remove
exit
  • You'll be able to see in Disk Management the Recovery (or whatever you named it) partition loses it's new drive letter

Optionally:

  • Confirm the recovery partition is working with reagentc /info
  • Boot into recovery reagentc /boottore /logfile C:\Temp\Reagent.log
  • Delete the recovery image file del C:\recovery-image.wim

Prevent Automatic Mounting of New Volumes

NOTE: this still needs tested, and does not appear to work as expected.

This is useful in forensics where you want to connect an external drive but do not want to mount or open the filesystem.

On Windows Server, this is a built in utility: automount.

On a Windows workstation, you must use mountvol.exe.

Disables automatic mounting of new basic volumes. New volumes are not mounted automatically when added to the system.

mountvol.exe /N
  • Now mountvol.exe /? will print a message at the bottom of the usage information noting if automatic mounting is disabled

Re-enables automatic mounting of new basic volumes.

mountvol.exe /E

EFI Partition

This mounts the \EFI partiton under E:\ if it's available (you can use any available drive letter).

mountvol.exe E: /S

If you want to review the 10 most recently modified files in the EFI partition mounted at E:\:

gci E:\EFI -Recurse -Force -File | sort LastWriteTime -Descending | Select -First 10 FullName,LastWriteTime 2>$nul

Get a list of all filetypes found in E:\EFI:

gci E:\EFI -Recurse -Force -File | sort LastWriteTime -Descending | Group-Object Extension 2>$nul

Check the signature of every file in E:\EFI, using C:\Tools\sigcheck64.exe:

foreach ($efi_file in (gci E:\ -Recurse -Force -File).FullName) { if (C:\Tools\sigcheck64.exe -accepteula $efi_file | Select-String "^\s+Verified:\s+Signed$") { Write-Host -ForegroundColor GREEN "[OK]$efi_file" } else { Write-Host -ForegroundColor RED "[WARNING]$efi_file" } }

Device Management

Restricting device and driver installation.

Device ID Strings

https://learn.microsoft.com/en-us/windows/client-management/manage-device-installation-with-group-policy#determine-device-identification-strings

The Class GUID is the type of device, for example a printer or a tablet.

The Instance ID, Hardware IDs, and Compatible IDs are identifiers for the device. The most specific is the Instance ID. This relates to that one device, and individual devices may have multiple Instance IDs. You'll need these ID's to 'allow' or 'deny' devices.

Enum Device ID Strings (cmd.exe)

You can start like this from a command prompt with pnputil:

pnputil /enum-devices /connected

If you were trying to find a single device, scroll through the list to determine the possible Class GUID of the device you're looking for.

Next you'd query connected devices of that class. This should narrow down the results enough for you to identify the exact device.

pnputil /enum-devices /connected /class '{<guid>}'

In this case, the Device Description fields most closely matching and describing your device are usually the right ones.

Finally, query the device directly using the Instance ID (there may be multiple Instand IDs for the same device, query each one you find) and this time also print the Hardware IDs and Compatible IDs for each instance:

pnputil /enum-devices /instanceid '<instance-id>' /ids

Enum Device ID Strings (PowerShell)

All the credit for this one goes to the PowerShell Team over on the Microsoft DevBlogs:

This will gather all of your connected USB devices in a single line:

Get-WmiObject Win32_USBControllerDevice | %{[wmi]($_.Dependent)} | Sort-Object Description,DeviceID | ft Description,DeviceID -auto

We can append ClassGuid, HardwareID, and or CompatibleID onto an fl instead of an ft command to obtain the other IDs as well:

Get-WmiObject Win32_USBControllerDevice | %{[wmi]($_.Dependent)} | Sort-Object Description,DeviceID | fl Description,DeviceID,ClassGuid,HardwareID,CompatibleID

In any case, you'll want to note the Instance ID, Hardware IDs, and Compatible IDs for use with policy configuration.

As mentioned in #3 of Scenario steps - preventing installation of prohibited devices, be sure you know exactly what will be blocked by the policy before deploying it. It's recommended to test all of this in a VM, as you can quickly recover from a policy that's too broad (which could block all external devices preventing you from using or recovering the machine).

Devices are evaluated by the Apply layered order of evaluation for Allow and Prevent device installation policies across all device match criteria policy in the following order:

  • Device Instance IDs
  • Device IDs (hardware / compatible)
  • Device Setup Class
  • Removeable Devices

Device Instance IDs are the most specific and always take precedence.

To allow only trusted devices while defaulting to blocking all others:

  • Enter the ID's of devices you wish to allow under Allow installation of devices that match any of these device instance IDs and enable this policy
  • Set Apply layered order of evaluation for Allow and Prevent device installation policies across all device match criteria policy to enable
  • Enabled the Prevent installation of removable devices policy

Alternatively, and as a test, if you wish to block specific devices while allowing all by default:

  • Enable the Prevent installation of devices that match any of these device instance IDs policy to block specific Instance IDs (precise)
  • Enable the Prevent installation of devices that match any of these device IDs policy to block based on Hardware or Compatible IDs (broad)
  • Click Show and add the IDs to this list
  • Check Also apply to matching devices that are already installed

What the last point will do is disconnect any currently connected device(s) matching the policy and remove their drivers. In Windows 11 this will happen immediately after applying the policy.

Remember you may need to specify multiple Instance IDs even for one device.

You can confirm the devices are no longer visible with:

pnputil.exe /enum-devices /connected /class '{<guid>}'

Disabling or setting that policy to Not Configured will reconnect the devices.

Deny Execute Access, Disable Autorun

By default, Windows will automatically mount any external drives connected. However, these settings will prevent them from executing any content. AutoPlay typically loads media or external files automatically based on a user setting. Autoruns (since Vista) execute content from an autorun.inf file. These files should require user interaction.

In the following GPO path:

Local Computer Policy > Administrative Templates > System > Removable Storage Access

The following policies control whether content can execute at all from removable media.

  • CD and DVD: Deny execute access
  • Floppy Drives: Deny execute access
  • Removable Disks: Deny execute access
  • All Available Storage Classes: Deny execute access
  • Tape Drives: Deny execute access

Next, in this GPO path:

Local Computer Policy > Administrative Templates > Windows Components > AutoPlay Policies

These settings control Autorun and AutoPlay features.

  • Turn off Autoplay: Enabled, CD-ROM and removable media drives
  • Set the default behaivor for Autorun: Enabled, Do not execute any autorun commands
  • Disallow Autoplay for non-volume devices: Enabled

To set these items from PowerShell:

# Per user, AutoPlay setting
Set-ItemProperty -Path "HKCU:\Software\Microsoft\Windows\CurrentVersion\Explorer\AutoplayHandlers" -Name "DisableAutoplay" -Type DWord -Value 1

Enable or Disable PnP Devices

stackoverflow: Enable / Disable Webcam with PowerShell

Sometimes you won't be able to re-enable a previously disabled webcam through the settings menu. Instead enumerate cameras, then enable or disable them using their InstanceId.

Enumerate known cameras.

  • Disabled cameras have a status of error
  • Enabled cameras have a status of OK
  • Disconnected cameras have a status of unknown
Get-PnpDevice -FriendlyName "*cam*" | select Status,FriendlyName,InstanceId | fl
Get-PnpDevice -Class camera | select Status,FriendlyName,InstanceId | fl

Enable / Disable camera as administrator, you'll be prompted to confirm:

Enable-PnpDevice -InstanceId "<instanceid>"

Blocking ISO Mounting

Taken directly from Mubix's blog post:

What this does:

  • Block ISO mounting from double clicking
  • Block ISO mounting via the context menu
  • Block programmatic ISO mounting via PowerShell

In the following GPO path:

Local Computer Policy > Administrative Templates > System > Device Installation > Device Installation Restrictions

Enable this policy:

Prevent installation of devices that match any of these device IDs

Check the "Also apply to matching devices that are already installed" option.

And set this as a value:

SCSI\CdRomMsft____Virtual_DVD-ROM_

Try to mount an ISO to validate this is working. All ISO files should fail to mount to the filesystem once this policy is in place.

Creating an ISO for Testing

On Ubuntu the genisoimage command can quickly create an ISO of any folder or file:

genisoimage -o <outfile>.iso <input-folder>
genisoimage -o my-docs.iso ./Documents

You can verify the ISO and contents with:

file ./my-docs.iso

sudo mkdir /mnt/my-iso
sudo mount my-docs.iso /mnt/my-iso

ls -l /mnt/my-iso

sudo umount /mnt/my-iso
sudo rm -rf /mnt/my-iso

Now move the .iso file over to your Windows machine (or if you did this in WSL it's already there) to confirm your policy configurations are working.

Additional Use Cases

TO DO: how to configure these policies with just PowerShell


About

Various configuration files for Microsoft Windows operating systems

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published