Writing
Notes on building and running infrastructure. Each piece describes a system I operate or a problem I solved. Technical deep dives are linked where the implementation detail warrants it.
An AI Browser Agent as a Low-Vision Computer Tool
I use ZoomText. An AI that reads and fills web pages in my own browser, while I watch and make the final click, ended the hunt for hamburger menus and three dots. How it works over the Chrome DevTools Protocol, what it did, and the rules that keep it safe.
How to Sandbox AI Agents on a $25 VPS
Two AI coders, Grok Build and Codex CLI, share one small server without clobbering each other: separate users, per-agent egress allowlists, a private Forgejo with read-only cross-review, and the five things that went wrong on the way.
Why I’m Getting the CKA
Why Kubernetes matters, why the CKA certification is worth pursuing, and what running a production K3s cluster with 14 services teaches you that tutorials cannot. A case for hands-on certification in the age of container orchestration.
ACE vs CAEC: Two Paths to Ontario High School Equivalency
Comparing Ontario's two adult high school equivalency credentials: the free, course-based ACE Certificate at Fanshawe and the $100 test-based CAEC at TVO ILC. Why running both in parallel is the strongest strategy for adult learners building toward OSAP-funded post-secondary education.
Learning Python Formally When You Already Live in the Terminal
If you already think in bash and vim, the gap is not skills—it is credentials. A free course and a $69 exam close that gap. Why now, and why Python is the natural next language for someone who already runs infrastructure from the command line.
Indexed Code Search for Gitea with MeiliSearch and MCP
Replacing sequential grep with sub-50ms indexed search across 38 Gitea repos. A Go indexer feeds MeiliSearch, exposed as an MCP tool for LLM agents. The effective searchable knowledge base grew from 4K lines to 993 documents while reducing always-loaded context by 18%.
Technical Writing
Implementation detail, reference material, and deep dives. These are longer and more specific than the posts above — written for someone who wants to understand or reproduce the system, or for an LLM agent retrieving context.
First deep dives coming soon.