blog/

leaving-windows-for-a-cachyos-hyprland-rice.md

Why I Left Windows and Built a CachyOS + Hyprland Desktop That's Actually Fast

I was tired of a 30GB OS eating background CPU just to sit idle, forcing updates on its own schedule, and fighting me every time I wanted to change how it looked. So I rebuilt my daily driver from scratch on CachyOS + Hyprland — a desktop I fully control, that stays under a gigabyte of overhead at idle, and that I never have to touch a mouse to use. Here's the whole build: the tools, the reasoning, and the receipts.

gpu-programming-why-it-matters-and-how-to-start.md

GPU Programming, Demystified: What It Is, Why Everyone Suddenly Needs It, and How to Actually Start

The whole tech world is fighting over GPUs right now — and most people still think they're just for gaming. This is the field guide I wish I'd had: what a GPU really is, why the AI era made GPU programming a superpower, and the exact path (hardware, tools, and a first working kernel) to write code that runs on thousands of cores at once.

netengine-open-source-network-toolkit.md

NetEngine Is Now Open Source: Building One Desktop Console to Replace Ten Network Tools

I got tired of juggling a scanner, a terminal, PuTTY, WinSCP, Postman, and a ping window just to fix one network problem. So I built NetEngine — eight network tools in one private, local-first PyQt6 console with an AI automation layer. Here's how it works, what I learned building it, and why it's now free and open source.

mcp-and-rag-the-new-automation-stack.md

MCP + RAG: The Stack That's Quietly Rewiring AI Automation in 2026

From a simple Obsidian plugin to full agentic pipelines that search, retrieve, and act — here is how the Model Context Protocol and Retrieval-Augmented Generation are being combined to build the next generation of AI automation tools.

something-new-is-coming.md

Something new is coming

A quick heads-up — a new post is in the works, and it lands here soon. This is a short note to say thanks for subscribing and that more is on the way.

anthropic-mythos-and-claude-opus-4-7.md

Anthropic's Mythos Moment: Inside Claude Opus 4.7 and the Model Too Powerful to Ship

Anthropic dropped two bombshells in April 2026 — a publicly available Opus 4.7 and a frontier 'Mythos' model so capable at finding zero-days that the company kept it private. Here's what that means for AI, security, and everyone building on top of it.

ai-in-cybersecurity-2026.md

AI in Cybersecurity in 2026: The Arms Race Between Attackers and Defenders

AI is now on both sides of every attack. From deepfake phishing to autonomous malware to agentic SOCs, here is what is actually happening in cybersecurity in 2026 — and what to do about it.