Since my agents are handling the merge conflicts, I’m fearlessly working on the same project on two machines in parallel 😎
I mean ... my agents do 😉
Code reviews worked for a long time.
They caught design smells
and even bugs.
But they are not a sustainable
quality strategy in the age of AI.
We need other mechanisms
to ensure both the internal
and external quality of our software. 🤔
Seb retweeted
People seem shocked that the author of Clean Code doesn't read code anymore. But that should not be surprising.
Why did I write Clean Code? What is the goal of keeping code clean? To get it out of the way of the real job -- thinking.
The longer the source files,
the slower the coding agents seem to get.
Do you see the same pattern in your codebase? 🤔
What is the biggest bottleneck
for your coding agents?
Context retrieval?
Model latency?
Test execution?
🤔
Luckily, the specific programming language becomes less and less relevant 👇😉
techinfo.pub/2026/04/10/NET-…
A lot of good design work used to be intuition:
“This does not feel right yet.”
Then we refactored until it did.
With coding agents,
that intuition has to become explicit:
detecting design smells
without reading every line of code,
and expressing exactly what is wrong
so the agents can fix it as intended.
Clearly one of the hardest parts
of agentic development. 🤔
Thrilled to announce that @ThePrimeagen has joined Omarchy Core to lead Agentic QA. His new Oligarchy harness will soon be vetting all our upcoming releases using @digitalocean droplets, which will host agents running on our @AIatMeta tokens. Full cricle! omarchy.org/news/2026/09/the…
"So you accept failure in the gym
because you know it’s optimal for growth,
but in every other aspect of life ... " 😉
And still - in my company - most devs write code by hand 🤷♂️
It's pencils down, people. Writing code by hand is no longer an economically viable skill for most programmers at most companies. But the future of making software has never been brighter. Don't you dare black pill this beautiful moment! youtu.be/vDjW_dRyKXY?si=6Fsf…
Sonnet is so chatty.
I found a class with >500 LoC -
more than half of it: code comments 🤦♂️
Haven’t seen GPT Luna Max do that yet 😉
Seb retweeted
A simple demo of uml-viewer. This is how I review the projects I create with agents nowadays.
This video is larger than Cloudflare's 512 MB cache, so it can't be played through. More donations are needed to cover a larger cache. Donate
Yesterday I switched my coder agent from Sonnet to Luna Max and tried it on a few features.
Similar speed.
Similar quality.
Significantly fewer tokens.
The only adjustment:
I asked my Architect to create even smaller slices.
I like this setup. Will keep it for a while. 😉
💯💯
Replying to @anah_sahh
I’m having more fun with software than at any time in the last decade. And the last decade was a lot of fun for me.