In the previous post we looked at hidden furnaces in agentic products. The example was the most common one, title generation. You type the first message, and the product quietly asks another model to name the conversation. Small and useful.
Something that is important to understand when working with agentic products is that we are always in the hands of the product developer.
The most interesting AI workflow I have right now is not a coding workflow. It has code in it. It has a small Swift CLI, some auth state, a few JSON exports, a Codex automation, and way too much knowledge about how Discourse represents unread things. But it is not one of those fancy loops that will make you a product overnight.
Since I’ve been working with AI agentic tools, and especially since I’ve been working on the various incarnations of my own agent harness, there is something that is very clear to me.
WWDC week has always been a funny mix of excitement and overload for me.
A few weeks ago I read An Interpreter for Swift, from Cocoanetics, and I had that nice feeling of somebody else having already articulated the thing that was vaguely floating in your head.
Skills are everywhere right now. People talk about them as if they were some magical incantation that will make AI more powerful while also keeping it under control.
In the previous post we gave superpowers to the harness in the form of a bash tool. Even doing it just for the learning experience, it was already very eye-opening. It made this little Swift harness feel much closer to the real thing.
After teaching the harness to load AGENTS.md, my mind went back to thinking about the tools, and wondered, do we even need multiple tools?
In the previous post, Have You Built an Agent Harness Yet?, we built a tiny harness in Swift. We gave it a loop, a prompt, and some tools, and with that it already had most of the features we are used to with modern AI. No magic. Just autocomplete, and a harness deciding what the model gets to see and do.
For years I have repeated a thing that I still believe. Every programmer should write a promise library once. I think agent harnesses are the 2026 version of that exercise.
I have this weird feeling of deja vu.