Security is ultimately a political problem that may have technical solutions. It's because we can always risk it and hope for the best. Observe a climber, solving the technical problem of ascending after a political decision of taking this route.
The internet is full of bots and some aren't nice (DDoS, scraping, exploits, AI..). To block bots, we must identify them in the traffic. That's hard as some pretend to be human/browsers! Let's explore JA3/JA4 hashes, fingerprinting the TLS handshake.
Internet traffic is mostly encrypted - one SNI extension still causes trouble. Encrypted Client Hello aims to fix this for a price of dependency on big tech. Is this the way?
There was a recent incident where Microsoft somehow allegedly blocked a mailbox of a sanctioned individual. Any organization highly depending on MS products that might come into the crosshair should ask - can this happen to me?
A brand new feature in LLMs is “understanding” text. Pretty much all of the knowledge work can be abstracted to a simple instruction: Answer my question using the relevant context. LLM hype claims it's solved. I’ll show you that it’s very much not.
In this part, we’ll be looking at Model Context Protocol (MCP). We’ll cover the features (tools, resources, prompts, roots, sampling), how it works, how to communicate with it (both locally and remotely).
I need to write a python script to perform/automate some task from time to time. But I don't like writing the same thing over and over again and I sure don't need a package. So I wrote a cookiecutter template for my scripts.
I got a message from my friend in the lines of: “Hi! I remember you did penetration tests, we want some. Where can we order some penetration tests?” Dear friend, here's your answer.