GitLab Path Traversal Vulnerability CVE-2026-85706 Exploitation Imminent

Added a network-layer detection project to my Blue Team portfolio: Snort IDS/IPS 📡 Built on top of my earlier SOC Detection Lab (Splunk) — same attack scenarios, different detection layer. This time I wrote custom Snort 3 rules to catch SSH brute-force attempts and port scans directly at the network level, then validated them against real attack traffic (Hydra + Nmap) from a Kali VM. Snort 3's Lua-based config threw some real curveballs along the way — a missing comma breaking the parser, VirtualBox checksum offloading silently dropping packets before rule matching, relative path resolution issues. Documented all of it, because the debugging is often the most useful part. End result: two custom rules, both firing correctly on live traffic, confirmed via console alert output. Full writeup on GitHub 👇 https://lnkd.in/gFbpP6sJ #CyberSecurity #SOC #BlueTeam #Snort #IDS #NetworkSecurity #InfoSec #OpenToWork

Day 3: TCP 3-Way Handshake Analysis & Stream Inspection 🛡️ Stepping up my SOC Analyst lab intensity! Today’s practical focused on transport layer analysis—specifically inspecting how TCP connections establish and stream payloads interact. Key Practical Steps Taken Today: Triggered web requests via CLI (curl) to observe raw TCP connection initialization. Applied tcp.flags.syn == 1 filters to isolate connection requests (SYN packets). Used Wireshark’s Follow TCP Stream feature to reconstruct and analyze full client-server communications. Understanding connection dynamics at the packet level is critical for identifying port scans and unauthorized network reconnaissance. On to Day 4! 🚀 #Cybersecurity #SOCAnalyst #Wireshark #NetworkSecurity #KaliLinux #LearningInPublic #WomenInTech

Here's the traffic flow I mapped out for my Wireshark lab - from the internet, through the network, into the interface, and finally into Wireshark for inspection. Understanding this flow is what makes packet analysis click. Once you can picture where your capture point actually sits in the path, everything else - filters, handshakes, DNS lookups - starts to make a lot more sense. A few things this diagram helped me internalize: - Wireshark only sees what reaches the interface you're capturing on - Promiscuous mode is what lets you see more than just your own traffic on a switched network - Every layer (frame, IP, TCP/UDP, application data) is visible and inspectable, one at a time If you're learning packet analysis, sketch this out yourself before touching the tool. It changes how you read every capture after. #CloudEngineering #AzureCloud #CloudSecurity #NetworkEngineering #Wireshark #CyberSecurity #ITCareers

NVIDIA PSIRT Addresses Critical Vulnerabilities in DGX Spark and OpenClaw: A Technical Deep Dive into Firmware Flaws and AI Agent Security + Video Introduction: NVIDIA’s Product Security Incident Response Team (PSIRT) has recently addressed a wave of vulnerabilities affecting two cornerstone products in the AI development ecosystem: the DGX Spark workstation and the OpenClaw/NemoClaw AI agent framework. While NVIDIA does not operate a traditional bug bounty program, its PSIRT has demonstrated a commitment to coordinated vulnerability disclosure, acknowledging researchers who report security issues under its policies....

This research highlights an important hardware security lesson. The exploit chain targeting Unisoc chipsets wasn't just about a VoLTE vulnerability. The bigger concern was the architectural trust boundary between the modem and the application processor. When components share memory without strong hardware isolation, a compromise in one domain can become a compromise of the entire device. This is why hardware security matters long before software defenses come into play. Security shouldn't rely solely on patching vulnerabilities. It should also be built on architectures that assume one component may eventually fail and prevent that failure from spreading. As AI, connected devices, and mobile platforms continue to evolve, hardware-rooted isolation, secure-by-design architectures, and coordinated vulnerability response will become even more important. Strong security starts with strong foundations. 🔐https://lnkd.in/g7FVttSG #Cybersecurity #HardwareSecurity #Android #EmbeddedSecurity #MobileSecurity #SemiconductorSecurity

The Offensive AI Shift and Orbital Cyber Risk: A Technical Deep Dive + Video Introduction: The cybersecurity landscape is undergoing a foundational transformation as legacy penetration testing tools face obsolescence against AI-driven autonomous exploit systems. The vertical is no longer about simple vulnerability scanning but about supervisor agents that orchestrate recon, execution, and dynamic payload rewriting, compressing months of work into hours. Coupled with the militarization of space where unencrypted satellite traffic and quantum decryption threats are now tangible risks, the technical community must integrate new toolchains and hardened configurations across Linux and Windows environments....

Wireshark Network Traffic Analysis: End to End As part of my hands-on networking lab work, I worked with Wireshark to analyze real network traffic and understand what is actually happening behind network communication. I created this beginner-friendly visual from my lab experience to help anyone who is still struggling to understand packets, protocols, IP addresses, ports, TCP connections, and the OSI model. The guide walks through the Wireshark interface, packet filtering, the TCP three-way handshake (SYN → SYN-ACK → ACK), common protocols, packet headers, and how data moves through different network layers. My lab also involved examining Ethernet, IPv4 and TCP headers to understand fields such as source/destination addresses, ports, sequence numbers and acknowledgment numbers. My goal in sharing this is simple: turn what I learned through hands-on practice into something that may make networking easier for someone else. #Wireshark #Networking #NetworkEngineering #NetworkAnalysis #Cybersecurity #CloudComputing #DevOps #LearningByDoing #HandsOnLearning

An embedded device is not simply a smaller server... Yet I often hear that we should secure embedded systems by applying the same techniques already used in IT. Some of them are certainly "compatible", like cryptography or authentication, but they cannot simply be copied without first considering the device architecture, available resources, performance requirements, and expected lifetime. A Linux-based device may support a sophisticated firewall, while a small MCU may have limited RAM and flash, strict real-time deadlines, no memory management unit, and may remain deployed for 15 years with only occasional updates. In that environment, security requires a different mindset! The focus may move toward minimizing the attack surface, establishing trust during boot or using hardware isolation.. Even familiar mechanisms such as TLS must be evaluated against memory usage, execution time and hardware acceleration. The objective is still the same: protect the system! But the engineering path depends heavily on the system we are protecting. Which IT security technique have you found most difficult to adapt to an embedded device? #EmbeddedSecurity #Cybersecurity #EmbeddedSystems #SecurityArchitecture #IoTSecurity

Bad actors have access to the best tools, we all know this. How quickly can you recover? More importantly have you validated RTOs/RPOs for scenarios like this?

Proud to be working with a team that's fighting the good fight. 💪 If we can understand how these attackers are exploiting tools like Cursor and AI models like Claude Sonnet, we can help businesses prevent attacks and also stay up in the event that they get hit. Amazing work, Gambit Security research team! Another great technical writeup by Eyal Sela! https://lnkd.in/dAWnNaGk