Reference

Glossary

The language of expression infrastructure.

D\ introduces a new vocabulary because it occupies a new category. These terms are not jargon. They are the precise names for ideas that did not exist before expression infrastructure did.

Core Concept

Taste

The accumulated weight of human judgment about what feels right. Not subjective preference. Not decoration. The ability to discern why one arrangement of elements feels considered and another feels careless.

Taste is what makes a typeface feel inevitable at a certain size. What makes spacing breathe instead of suffocate. What turns a functional interface into one that feels made for humans.

Why it matters: Taste is the dimension machines are missing. It is what separates functional from felt. D\ exists to encode it as infrastructure.
Core Concept

Expression Infrastructure

The foundational layer that makes design judgment available to machines as a runtime service. Not a design tool. Not a component library. Infrastructure – like an API, but for taste.

Three quadrants of interface tooling were full: human + templated (page builders), human + composable (design systems), agent + templated (AI builders). D\ fills the fourth: agent + composable.

Why it matters: The category did not exist before D\ because the problem did not exist before coding agents. Now it does. And it needs infrastructure, not tools.
Core Concept

Expression Layer

The layer of any interface where design decisions happen. Color, typography, spacing, rhythm, hierarchy – every surface has an expression layer, whether someone designed it intentionally or a machine generated it by default.

In traditional software, this layer is manual. In the agentic era, it is automated through expression infrastructure.

Why it matters: Every interface has an expression layer, whether intentional or accidental. D\ makes it intentional.
The Name

D\

Design, less. The backslash is an escape character – a departure from the old paradigm where design required designers at every step. Read “design-less”: less struggle, less soullessness, less of the old way.

D is the half-circle serif – design, intention, craft. The backslash is the exit. Together they say: there is a way out of the world where machines build without taste.

Why it matters: The name is the thesis. The best design is the design you do not have to fight for.
Artifact

Brand Capsule

A self-contained compilation of everything that defines how a brand should look, feel, and speak. One artifact. Versioned. Deterministic. When agents consume a capsule, they produce brand-consistent output automatically.

Think of it as a brand’s DNA – compressed into a single, immutable package that any machine can read and any human can trust.

Why it matters: Without it, every agent interaction starts from zero. With it, brand consistency becomes infrastructure – not oversight.
Artifact

Brand Profile

The living representation of a brand’s design preferences. Not a static guideline document. A profile that refines with every interaction, building an increasingly precise understanding of what the brand should feel like.

A Brand Profile is to expression what a user profile is to personalization – the accumulated knowledge that makes every next interaction better.

Why it matters: Static brand guidelines go stale the day they are published. A Brand Profile evolves with the brand.
System

Brand Operating System

The governing layer that compiles, validates, and distributes a brand’s expression identity. It ensures every surface an agent touches speaks with the same intention, the same voice, the same considered judgment.

Where a brand guideline is a document people read, a Brand Operating System is infrastructure machines run.

Why it matters: Brands are expressed across hundreds of surfaces by dozens of agents. A Brand OS ensures coherence without manual oversight at every step.
Agent

Designless Agent

The agent at the center of Designless. It works inside the AI tools you already use and orchestrates the work of expressing, building and serving in your brand, reasoning with the Designless Language Model and your Brand Capsule.

The Designless Agent works with named sub-agents: Prism paints the canvas, Arbiter governs the brand, and LoomX supports and answers.

Why it matters: The agents you use already build fast. The Designless Agent lets them express with the design and brand intelligence you already have, so what they build carries it.
Sub-agent

LoomX™

The support and answer sub-agent of the Designless Agent. LoomX answers questions about Designless and helps people get set up.

It works beside Prism, which paints the canvas, and Arbiter, which governs the brand. The Designless Agent does the work. LoomX makes sure the people using it are never left stuck.

Why it matters: New infrastructure raises questions. LoomX answers them in plain language, so the path from a first question to a first build stays short.
Sub-agent

Prism

The sub-agent of the Designless Agent that paints the canvas. When you ask for something visual, such as a deck, a carousel, a poster or a page, Prism renders it on the Designless canvas in your brand.

The work lands where you and your team can see it, review it and edit it, instead of arriving as code or text to imagine.

Why it matters: Design is judged by looking. Prism puts every artefact in front of the people who decide, in the brand it was made for, so judgment happens on the real thing.
Sub-agent

Arbiter

The sub-agent of the Designless Agent that governs the brand. Arbiter checks what is created against your brand’s rules before it reaches you, fixes what it can, and flags what needs a person’s judgment.

Every artefact answers to the same rules, whichever agent made it and whoever asked for it.

Why it matters: A brand erodes one small exception at a time. Arbiter holds every artefact to the brand’s rules and leaves the real judgment calls to people.
Engine

Designless Language Model

The engine of Designless. The Designless Language Model, or DLM, does two things. It reads intent: it understands the design intention behind a request, what is being asked for and why, so the work starts from the right question. And it carries judgment: the compounded sum of fractional design decisions, the judgments designers make that machines do not know.

That judgment is instrumented, not written down. Every accepted decision is a measurement and every override is a correction, so taste stops being a claim and becomes a record. It is the design judgment the Designless Agent reasons with. The model that makes taste computable.

Why it matters: Agents know logic. They do not know what a person means when they ask for something designed, or the many small decisions that make a brand feel like itself. The Designless Language Model is where both live.
Engine

LESS Studio

The workspace where brands create, refine, and manage their expression systems. Where human taste gets encoded into infrastructure agents can use.

LESS Studio is where the authoring happens – the place where a brand’s identity becomes something machines can invoke.

Why it matters: Expression infrastructure needs a place to be authored. LESS Studio is that place.
No terms match your search.