Emil Kowalski Thinks Taste Is Teachable. His Skill Is the Argument.
The Emil Kowalski Design Engineering skill packages his micro-interaction sensibility as agent instructions. The claim it makes — that taste is encodable in repeatable patterns — is a bold one worth examining.
If you've spent time on X following design engineers, you've seen Emil Kowalski's work. The spring animations that feel physically right. The hover states that have just enough movement to signal interactivity without demanding attention. The transitions that make navigation feel like space rather than reloading.
His work has a specific quality that people describe as taste — an aesthetic sensibility that seems hard to teach, harder to specify, and impossible to reduce to rules.
He put it in a skill anyway.
What Micro-Interactions Actually Are
Before getting to what the skill does, it's worth being specific about what micro-interactions are, because the term is used loosely.
Micro-interactions are the small, moment-to-moment responses a UI gives to user actions. A button that depresses slightly when clicked. A text field that shakes when validation fails. A list item that glides away when deleted rather than disappearing abruptly. A menu that slides in from the correct direction based on where the trigger is.
These are not features. They don't add functionality. They add felt quality — the difference between an interface that feels cheap and one that feels crafted.
Emil Kowalski's reputation is built on doing these well — specifically, on using spring-based animations and physics-based timing that makes interfaces feel like they have weight and momentum rather than duration.
What the Skill Encodes
The Emil Kowalski Design Engineering skill from emilkowalski/skill has 29,400 installs. The skill packages his micro-interaction approach as agent instructions — specific guidance on when to use which animation properties, what spring configurations produce which felt qualities, when motion should be reduced or eliminated.
What's interesting is that the skill doesn't describe general animation principles. It describes Emil's specific aesthetic choices. Not "use spring animations for natural feel" but the actual parameters and timing decisions that distinguish his work from generic motion design.
That specificity is unusual. Most design skills are principles. This one is an individual's aesthetic encoded as instructions.
The "Taste Is Teachable" Argument
Here's the claim the skill implicitly makes: what makes Emil Kowalski's interfaces feel the way they feel is not some ineffable creative quality. It's a set of specific, repeatable decisions that can be described precisely enough to follow consistently.
That's a bold claim, and I think it's partially right.
The aesthetic sensibility — the judgment about when to add motion and when restraint is better — is genuinely hard to encode. The skill can't replicate the years of taste formation that went into deciding which interactions deserve animation and which ones would become noise.
But the execution layer — given that a particular interaction should have a spring animation, what should the stiffness, damping, and mass be to achieve the right felt quality — that is encodable. Those are numeric decisions. They have better and worse values. Emil has been making them correctly for long enough that his values constitute a calibrated library.
What the skill provides is calibrated defaults for the execution layer. If you're using it and you have good judgment about when interactions need motion, you'll produce work that feels like Emil's. If you're using it without that judgment, you'll produce consistent motion design that may be inappropriately applied but will at least be well-executed.
What This Means for Design Engineering
The field of design engineering is young and its practitioners are still figuring out what's teachable versus what's taste. The existence of this skill is a data point in that conversation.
Emil Kowalski chose to package his approach as transferable knowledge rather than keeping it as personal craft. That's a statement about what he believes about his own work — that the patterns are the point, not the mystery.
I think he's right. The highest-value part of what design engineers do is knowing which interactions matter and having the conviction to invest in them when everyone else is treating UI as a secondary concern. That's judgment. That's hard to teach.
The execution of those interactions well? That's learnable. The skill is the evidence.
Part of the AI Skill Daily series — skills worth understanding, one at a time.