Every SOC automation platform has a deployment phase. Most have an optimization phase. Almost none have a learning phase.

Date: 22/04/2026

Explore jailbreaking techniques to bypass an AI model’s safety and policy restrictions.

How attackers bypass AI safety and break model restrictions

How attackers manipulate prompts to break AI system instructions

Server-Side Request Forgery (SSRF) is one of the most common and obscure vulnerabilities in the sphere of web security. Whereas most of the…

Anthropic is investigating a vendor breach after a Discord-linked group accessed its Claude Mythos AI model, with no evidence of impact on core systems.

If you’re at $0, you don’t need more knowledge. You need one working method and repetition.

API security When it comes to API Security and the attack surface our focus shifts from endpoints /login, /profile and the likes to objects, actions and trust assumptions within the API call An API …

Why Changing Your Password Doesn’t Always Evict an Attacker

One misplaced slash turned a file reader into a full filesystem browser

From data poisoning via anonymous S3 bucket to package hallucination supply chain attacks — here’s what conference research shows actually…

The world’s most widely used messaging platform is stepping into new territory. WhatsApp long known for its completely free core experience…

AirTouch HTB Writeup: Transparent Pivoting Without proxychains

Build SupplySecLab and learn to inspect models, audit dependencies, and govern LLM supply chains.

The premise of this challenge is to find 3 ingredients hidden in the system for a potion

In this writeup we solve the Dev Diaries room from TryHackMe using simple OSINT techniques The challenge shows how old snapshots public…

Understanding the Flow

“The boring findings pay the bills. The creative chains win the hall of fame.”