Pod Somm: Tracking Specialty Coffee with a Bookmarklet and a Lot of DOM Wrangling
How I built a personal coffee tracker for my xbloom studio, and why the interesting part was a bookmarklet that had to outsmart Cloudflare.
How I built a personal coffee tracker for my xbloom studio, and why the interesting part was a bookmarklet that had to outsmart Cloudflare.
How a script scaffolding tool grew from an interactive wizard into a full GitOps workflow, and what I learned presenting it at the Concord Software Developers meetup.
How I built a structured catalog of AI-generated cocktail recipes, and why the interesting part wasn’t the cocktails.
I joined Austin Lutz on the LutzTalk Podcast to talk about the shift to macOS GitOps, the creative side of IT engineering, Shikomi, AI, and what it takes to stay relevant in a changing industry.
As AI tools proliferate faster than security reviews can keep up, IT teams need fleet-wide visibility into what’s actually running. Three Jamf Extension Attributes for detecting LiteLLM, compromised npm packages, and AI gateway tools across your managed endpoints.
How I built a static Hugo site to replace a patchwork of Letterboxd lists and spreadsheets for a film club that’s been meeting since 2011, and what it took to turn 14 years of emails into structured data.
Why I built a lightweight native Swift app to preview Markdown and HTML files from the menu bar, and what I learned writing a custom Markdown renderer without any dependencies.
How I built a multi-source daily review system that pulls from Linear, Freshservice, my calendar, git history, and session diaries, and uses Claude to synthesize it all into morning plans, end-of-day summaries, and weekly retrospectives.
How I built a TestFlight beta tester management tool that stores everything (including the private key) in 1Password, implements JWT authentication from scratch in bash, and grew from a simple CLI into a native SwiftDialog GUI app.
How I built a two-component macOS update management system that ties CVE severity directly to deployment urgency, so users get accurate security context and admins get phased rollouts that automatically accelerate when it matters.