Recent content on XPN Infosec Blog
This post explores how effective LLM's are at reverse engineering EDR and AV engines. We look at Cortex, and just how GPT 5.5-Cyber was able to carve through each layer, providing YARA rules, Models, and behavioral detection rules to analyse.
In this post I walk through each major milestone that has been taken in my quest to explore building LLM-generated Mythic agents from prompt to deployment. Starting at the beginning, I’ll show what worked, what didn’t, and where this space is likely going next.
One of the things that has been winding me up about working with LLMs is how unmeasurable prompt modifications can be. You know how it goes, you task your agent to perform a task, and eventually it starts to deviate from the objective. So you head into your AGENTS.md to correct course, and hope that by shouting at the AI in just the right incantation, you fix the problem. GEPA is one established…
This article demonstrates how to jailbreak local models using constrained decoding.
In this article, we're going to look at how we benchmark the performance of our agent, and improve this capability with the macOS 27 Evaluations framework.
This article shows the power of macOS 27's new foundation model and OCR model for extracting credentials.
Dev Tunnels aren’t “just port forwarding”. They consist of layers of embedded protocols with RPC messages being exchanged. Once you peel the layers, you quickly see how Dev Tunnels are a C2 framework with extra steps.
A deep dive into discovering CVE-2025-64755, a vulnerability in Claude Code v2.0.25. This post walks through the process of reversing the obfuscated Claude Code JavaScript, and exploiting weak regex expressions to achieve code execution unprompted.
Microsoft will be introducing Administrator Protection into Windows 11, so I wanted to have an understanding of how this technology works and how it interacts with existing offensive tooling. While this technology is just a thin wrapper around a separate account, there are a few nuances such as who is permitted to access these accounts, as well as existing UAC bypasses which are still effective…
In this post we look at the new Prompt Guard 2 model from Meta, and introduce a concept I've been calling "Tokenization Confusion" which aims to confuse Unigram tokenization into generating tokens which will result in the misclassification of malicious prompts. We'll also look at why building up our ML knowledge will lead to better findings when assessing LLM API’s, as I discovered during a flight…
One of the things that I love about my role at SpecterOps is getting to dig into various technologies and seeing the resulting research being used in real-time. This post will explore one such story of how I was able to go from a simple request of recovering credentials from a database backup, to reverse engineering how SQL Server encryption works, finding some new methods of brute-forcing…
In this post we’re going to focus on some ADFS internals. We’ll be looking at OAuth2, and how it underpins the analogues to Entra ID security features like Device Registration and Primary Refresh Tokens.
Originally presented at SOCON-2024, and continuing the series into post-exploitation techniques against Identity Providers, in this blog post we'll look at Ping, OneLogin and Entra ID. I'll discuss how post-exploitation techniques effective against Okta apply to other providers, release new tools for post-exploitation, and look at what proves to be effective when critical assets lie beyond an…
While looking for avenues of injecting code into platform binaries back in macOS Monterey, I was able to identify a vulnerability which allowed the hijacking of Apple application entitlements. Recently I decided to revisit this vulnerability after a long time of trying to have it patched, and was surprised to see that it still works. There are some caveats introduced with later versions of macOS…
In this blog post, I'll discuss some of the post-exploitation techniques that I've found to be useful against Okta. Specifically, this post will look at how to use delegated authentication to our advantage, silver tickets, Okta AD agent spoofing, and finally how to deploy a fake SAML provider.
This year, LAPS 2.0 was released by Microsoft, and thankfully it now comes built-in to Windows. This time it comes ready for use with Active Directory, as well as being supported in Azure AD aka Entra ID. In this post, we’ll look at how LAPS 2.0 for Active Directory works under the hood, so you can make those fresh recommendations to your clients, and prepare yourself for the inevitable…
Steganography is experiencing a revival as a wrapper for delivering payloads. In this post we'll go back to basics and show just how steganography can be applied to a PNG image using the common least significant bit (LSB) encoding technique. No magic... just raw information... and a little C++.
In this blog we'll look at what it takes to construct an in-memory loader for Mach-O bundles within MacOS Ventura without using dyld. We'll walk though the lower-level details of what makes up a Mach-O file, how dyld processes load commands to map areas into memory, and how we can emulate this to avoid writing payloads to disk.
Up until recently, we've enjoyed in-memory loading of Mach-O bundles courtesy of dyld and its NSCreateObjectFileImageFromMemory/NSLinkModule API methods. And while these methods still exist today, there is a key difference.. memory modules are now persisted to disk. So in this post we'll take a look at just what was changed in dyld, and see what we can do to restore this functionality... hopefully…
This post looks at the recent trend of pulling Azure tokens from Office process memory and tries to identify just how these tokens were loaded, how Office handles a Microsoft Account (MSA), and how we can recover cached credentials from the Token Broker Cache.
In this post we'll explore just how SCCM uses its HTTP API to initialise a client, take a look at how Network Access Accounts are retrieved from SCCM, and see how we can decrypt these credentials without having to go anywhere near DPAPI.
With the leaking of code signing certificates and exploits for vulnerable drivers becoming common occurrences, adversaries are adopting the kernel as their new playground. And with Microsoft making technologies like Virtualization Based Security (VBS) and Hypervisor Code Integrity (HVCI) available, I wanted to take some time to understand just how vulnerable endpoints are when faced with an…
In this post, we'll dig into just how SMB over QUIC works, answer some of the immediate questions around which attacks are feasible, and show how we can repurpose some existing tooling to capture NTLM handshakes.
In this post we are going to look at one such technique that I thought was cool while playing around with the Windows Object Manager, and which should allow us to load an arbitrary DLL of our creation into a Windows process during initial execution, something that I've been calling "Object Overloading" for reasons which will hopefully become apparent in this post.
Recently I've been looking at the .NET CLR internals and wanted to understand what further techniques may be available for executing unmanaged code from the managed runtime. This post contains a snipped of some of the weird techniques that I found.
In this post, we are going to look at the Application Proxy protocol, how it works, and show how we can recreate enough functionality to allow us to create a custom inbound proxy into a client environment for our C2 traffic.
We've all been there, you've completed your initial recon, sent in your emails to gather those leaked HTTP headers, spent an age configuring your malleable profile to be just right, set up your CDNs and spun up your redirectors. Then it's time, you send in your email aaaaaand... nothing.
For a while I've wanted to explore the concept of leveraging a virtual machine on target during an engagement. The thought of having implant logic self-contained and running under a different OS to the base seems pretty interesting. But more so, I've been curious as to just how far traditional AV and EDR can go to detect malicious activity when running from a different virtual environment. While…
While working on some tooling recently I revisited the topic of .NET unmanaged exports and wanted to know just why this works in the way that it does. After all, by now we've all seen the COM calls required to spin up the CLR, so what makes unmanaged exports so special?
If you've never heard of MACL on MacOS, you're not alone. This obscure feature is a hidden part of MacOS that underpins Apple's concept of User-Intent, a shift in focus for MacOS privacy controls in an attempt to stop endless prompts interrupting the user. And by now we all understand just how annoying these alerts can be to us attackers. Being able to operate on an endpoint without giving the…
In this post, we are going to take a look at a couple of interesting methods of leveraging third-party technologies to achieve our code injection goals. For us, this translates to running code in the context of a target application without having to resort to disabling SIP.
.NET for post-exploitation is here to stay. It has been bundled with most C2 frameworks, common tools have been ported, AMSI has been added (then bypassed) and new and clever ways have been found to launch unmanaged code. The process of loading a .NET assembly however appears to be pretty consistent. We know that tools like Cobalt Strike's execute-assembly have greatly increased the accessibility…
It turns out that there is a method of disabling ETW in .NET, strangely exposed by setting an environment variable of COMPlus_ETWEnabled=0. This post explores how this works.
In this post we will walk you through the technology used to create and deploy the ActiveBreach Adversary Simulation Lab, and look at the hurdles we jumped through to get things running smoothly.
In this post we will focus on Event Threading for Windows (ETW), how it is used to surface events on .NET assemblies, and how we can evade this kind of detection.
In this post we will be looking at AWS Lambda, how we can deploy code using the Serverless framework, and how we can use this technology as a redirector for Cobalt Strike
As RedTeaming has grown with the industry, so has our need to build dependable environments. In keeping with the cat-and-mouse game we find ourselves in, it's essential to possess the capability of maintaining robust infrastructure which can be recreated if discovered, and more importantly, we need to ensure that the environment is free of issues upon deployment. Today I'm kicking off the first of…
Last year I posted a few tricks to help when targeting MacOS users, and included a technique useful for spoofing file extensions with the aim of taking advantage of Finder's removal of the .app extension from certain filenames.
In Cobalt Strike, blockdlls was introduced to allow protection of spawned processes from non-Microsoft signed DLL's. In this post I will show just how this works, and look at an additional process security option which could help us to deter endpoint security products.
Encountering Apple devices during RedTeam engagements is becoming increasingly common, so it's useful to have a few techniques available when navigating around whatever privacy or security changes are introduced with each version of MacOS. When MacOS Mojave rolled out at the end of 2018, a set of privacy restrictions were introduced to alert a user when an application requested access to sensitive…
In this blog post we will look at a somewhat familiar, but extremely limited window of opportunity which may come in handy when reviewing a fresh Active Directory forest deployment.
Hunting for new lateral movement techniques or interesting ways to execute code can be a nice way to sink some free time. With Windows spawning numerous RPC services on boot, finding unusual execution techniques is sometimes as simple as scratching just below the surface. And often the payoff far outweighs the time to discovery, with SOC or EDR vendors focusing on the more common published…
In a recent update to Sysmon, a new feature was introduced allowing the ability to log DNS events. While this gives an excellent datapoint for defenders (shout out to the SysInternals team for continuing to provide and support these awesome tools for free), for us as attackers, this means that should our implant or payloads attempt to communicate via DNS, BlueTeam have a potential way to pick up…
If you haven't had a chance to check it out, take a look here. Continuing on, in this post we will review what has become a nice way of subverting security controls added by Microsoft to prevent dumping of credentials, as well as extracting credentials as they are provided by a victim.
We’ve packed it, we’ve wrapped it, we’ve injected it and powershell’d it, and now we've settled on feeding it a memory dump, and still Mimikatz remains the tool of choice when extracting credentials from lsass on Windows systems. Of course this is due to the fact that with each new security control introduced by Microsoft, GentilKiwi always has a trick or two up his sleeve. If you have ever looked…
In this blog post we will explore some of our findings that might assist red teamers operating in environments where CylancePROTECT is in place and briefly touch on CylanceOPTICS, a complementary solution that provides rule based detection to the endpoint.
In this post I will be showing how to build a Azure DevOps pipeline for .NET projects, and hopefully show some techniques which I have found useful to modify build artifacts to make them a bit different, and in some cases, to increase the time it takes to analyse our tools if detected by Blue Team.
With clients increasingly relying on cloud services from Azure, one of the technologies that has been my radar for a while is Azure AD. For those who have not had the opportunity to work with this, the concept is simple, by extending authentication beyond on-prem Active Directory, users can authenticate with their AD credentials against Microsoft services such as Azure, Office365, Sharepoint, and…
In Cobalt Strike 3.13, the argue command was introduced as a way of taking advantage of argument spoofing. I was first made aware of the concept while watching Will Burgess's awesome talk RedTeaming in the EDR Age, with Will crediting Casey Smith who presented the idea during a series of tweets. As with anything introduced to Cobalt Strike which has the chance to improve operational security, I…
I’m not actually sure when the abstraction of Blockchain started or when it became such a marketing buzzword, but with so many things claiming to be “Powered by The Blockchain”, I wanted to dig into the technology to understand if there was any benefit to be had by an aggressor.