Blog posts
Deep dives into vulnerabilities we discover, offensive security techniques we develop, and insights from the front lines of enterprise security research.

LLM Security: Prompt Injection Defense with CaMeL Framework
DeepMind released CaMeL, a new prompt injection security mechanism. It is based on Dual LLM design pattern. Is it an end to prompt injection?
AI
LLM
Mateusz Wojciechowski
February 12, 2026
7
min read
•
Jun 30, 2025

GraphQL Security from a Pentester’s Perspective
Discover key security vulnerabilities in GraphQL APIs, practical testing methods, and effective mitigations. Essential reading for security professionals and developers.
API
Secure coding
Vulnerability research
Web
Paweł Zdunek
February 12, 2026
32
min read
•
Jul 7, 2025

DLL Hijacking in Check Point SmartConsole installer aka CVE-2024-24916
The article provides a detailed breakdown of a DLL hijacking vulnerability in Check Point SmartConsole, explaining how malicious DLLs could be loaded to achieve remote code execution before the issue was patched.
Binary exploitation
Red teaming
Vulnerability research
Karol Mazurek
February 12, 2026
5
min read
•
Jul 21, 2025

Race Condition Vulnerability Triggers Stack Buffer Overflow in macOS
macOS kernel vulnerability where a race condition between the display reset process and frame buffer updates, causing Stack Based Buffer Overflow which leads to kernel panic (potential privilege escalation) on Apple M1 systems.
Binary exploitation
MacOS
Reverse engineering
Vulnerability research
Karol Mazurek
February 12, 2026
12
min read
•
Aug 11, 2025

Visual Studio Code Vulnerability: macOS TCC Bypass
Microsoft Visual Studio Code on macOS has a misconfiguration in its Electron setup that enables the “RunAsNode” fuse, letting attackers bypass Apple’s TCC privacy protections and inherit sensitive permissions. It’s a proven flaw that works post-compromise, but Microsoft has refused to patch it, claiming it requires local access and doesn’t meet their servicing bar. The result is that any malware on the system could silently leverage VS Code to access private data without additional prompts.
Apple
MacOS
Red teaming
Vulnerability research
Karol Mazurek
February 27, 2026
3
min read
•
Aug 18, 2025

Zero Day Vulnerability Microsoft Delivered to macOS
Desktop application security lacked unified standards—until now. DASVS provides a structured approach to securing Windows, macOS, and Linux applications with clear verification levels and technical security controls. Our roadmap includes the Desktop Application Security Testing Guide (DASTG) and an automated security assessment tool. Join the community and help shape the future of desktop security!
Apple
Binary exploitation
MacOS
Vulnerability research
Karol Mazurek
February 12, 2026
7
min read
•
Sep 15, 2025

Threats of Unvalidated XPC Clients on macOS
The article discusses vulnerabilities in the Sparkle framework on macOS, related to improper XPC client validation. This leads to two main threats: a TCC Bypass (CVE-2025-10015) and Local Privilege Escalation (LPE) (CVE-2025-10016). Malicious applications can exploit these vulnerabilities to access TCC-protected files and escalate privileges. A proof of concept shows how an attacker can read sensitive files on the Desktop without triggering permission requests, highlighting the risks associated with services in applications like Ghostty.
Binary exploitation
MacOS
Red teaming
Vulnerability research
Karol Mazurek
February 12, 2026
15
min read
•
Oct 27, 2025

Mac Extended Attributes Expose Authentication Tokens in macOS
macOS stored complete download URLs – including authentication tokens and API keys – in persistent file metadata through the kMDItemWhereFroms extended attribute, exposing sensitive credentials to local attackers and malicious applications for lateral movement. Apple silently patched the vulnerability around macOS Tahoe without CVE assignment, public acknowledgment, or researcher credit, violating responsible disclosure norms two years after the initial report.
Vulnerability research
MacOS
Red teaming
Web
Karol Mazurek
February 12, 2026
3
min read
•
Nov 21, 2025

Format String Vulnerability in Apple’s TCC Daemon: A Deep-Dive
This post shows how an apparent format string vulnerability in Apple’s TCC daemon isn’t actually exploitable due to assembly-level stack manipulation that inadvertently provides the missing arguments. It demonstrates how bugs aren’t always what they seem, as low-level calling conventions, compiler optimizations, or decompiler artifacts can mask or fix high-level code issues.
Apple
Binary exploitation
MacOS
Reverse engineering
Vulnerability research
Karol Mazurek
February 13, 2026
6
min read
•
Aug 25, 2025
Previous
6 / 6
Sort by Categories
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Monthly Security Report
Subscribe to our Enterprise Security Report. Every month, we share what we're discovering in enterprise software, what vulnerabilities you should watch for, and the security trends we're seeing from our offensive security work.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Cookie Settings
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you.
See our Privacy PolicyThank you! Your submission has been received!
Oops! Something went wrong while submitting the form.