I am building ghx, a code reconnaissance CLI for AI agents, and now an agent sidecar framework around it. The hard question is whether any of it actually improves signal per token compared to plain old gh.

For a decade I believed great engineering would make everything else happen automatically. I am starting to understand how toxic that was.

SA_ONSTACK is process-wide. sigaltstack is per-thread. When Go, Bun, .NET, and the JVM share a process, signals land on the wrong stack. One kernel field fixes the class.

On realizing how easy it is to like things, want things, even feel things, and still keep them safely in the background.

Codemap, Aider, Gitingest, Repomix, and ghx solve different parts of agent code context: mapping, searching, packing, and deciding what to read before cloning.

How the Astral acquisition reveals the real war being fought in AI coding — and why everyone is talking about the wrong thing.

Every article you've ever read told you procrastination is the enemy. A flaw to fix. A habit to break. I spent 30 years believing that. Turns out, I had it completely backwards.

On Kat Zhang's obsidian note network, the GitHub ecosystem around it, and what graph neuroscience says about the shape of minds

A CLI tool that looks simple took 23 agent sessions, 2,500+ turns, and three rewrites. This is what building with AI actually looks like.

A new series: I find a source, go through it, distill what's worth your time, and tell you what I actually think. First up — Vercel's Open Source Program Winter 2026 cohort.

Software engineering isn't going away — writing code without agents is. The role is expanding, not shrinking. Here's what that actually means.