DESIGN SYSTEM / R&D

CHANDRA GARRE · SELECTED WORK

Weave

Weave

Weave

I gave four AI agents the same prompt. The two with no system agreed on 1 colour. The two with Weave agreed on 21. Then I asked one to audit my docs. It found a feature I’d promised and never built.

THE WORKING SYSTEM

Open Storybook

ROLE

Product Designer, solo

TYPE

Design system R&D

TIMELINE

Sep - Oct 2026

STATUS

35 components, 8 known gaps open

TOOLS

Figma, Storybook, React + MUI, Tailwind, Axe-core

At work I maintain enterprise design systems that ship to real users. You don’t test new token architecture on those. Weave is where I test it.

A laptop showing the Weave Storybook open on the Button page, with its table of live controls

AI stopped making bad screens. It started making someone else’s.

AI stopped making bad screens. It started making someone else’s.

Two runs of the same prompt picked the same arbitrary indigo and the same product name. I chose neither.

The model isn’t the problem. It has nothing to build from, so it decides for itself, then decides again next session. Then someone rebuilds the screen by hand before it can ship.

Corner of no-system run 1: an indigo logo and the product name Chalkline

NO SYSTEM, RUN 1

Corner of no-system run 2: an indigo logo and the product name Chalkline

NO SYSTEM, RUN 2

The same dashboard built with Weave components and tokensIllustration of a dashboard with no design system, made for comparisonNo systemWeave

Left: an illustration I made of a screen where nothing is decided for it. Right: the same screen built with Weave. The real, unedited agent runs are in Results.

Before Weave

After Weave

Developers
Before Weave

Rebuild each component by eye, from a screenshot.

After Weave

Open a component, set its props with live controls, and use it.

Designers
Before Weave

Decide in Figma, hand over a file, and find out later what was built.

After Weave

Decide in Figma, then check the real component in Storybook.

Agents
Before Weave

Guess, and guess differently next session.

After Weave

Read the rules first, then build with what exists.

35 components
8 OKLCH colour ramps
253 stories
506 passing renders, light and dark
  • Figma

  • Storybook 10

  • Vite

  • React + MUI

  • Tailwind 4

  • Axe-core

  • OKLCH

  • WCAG 2.1 AA

"weave.whychandra.in for developers. AGENTS.md for agents."

"weave.whychandra.in for developers. AGENTS.md for agents."

"weave.whychandra.in for developers. AGENTS.md for agents."

Figma is where I make decisions. But an engineer can’t import a decision, and an agent can’t read one at all.

Figma is where I make decisions. But an engineer can’t import a decision, and an agent can’t read one at all.

01/ 03

It can point at code. It can't run it.

Dev Mode has a component playground where you can flip props. The code panel shows generated code, or a snippet someone mapped by hand with Code Connect. Either way it's text. The real component lives somewhere else, so engineers still go looking for it.

Figma, Dev Mode
Button
width: 96px;height: 40px;background: #006B73;border-radius: 8px;
What it looks like.
Storybook
Button
<Button variant="solid" tone="primary" size="md"/>
What you type.
02/ 03

A drawn state isn't a working state.

Figma shows what each state looks like. Storybook shows how it behaves: focus ring, key handling, a name for screen readers.

Figma
ButtondefaultButtonhoverButtonfocusButtondisabled• • •loadingButtonerror
What each state looks like.
Storybook
Button
TabEnterEsc
Each one is a story you can open.
03/ 03

Two copies drift.

Design and code are two copies of the truth, and copies drift. Mine did. The docs said 44px tap areas. The code drew 32.

My docs
44
44px tap area around a 32px control.
My code
32
32px, and nothing around it.

Figma is still where I decide things. It isn't where they become true.

Four agent runs, one dashboard prompt. Two runs had nothing to read. Two had Weave.

Four agent runs, one dashboard prompt. Two runs had nothing to read. Two had Weave.

Run for run, the no-system screens looked fine. They just agreed on almost nothing. Two runs each is a small test, but the gap isn’t subtle.

Nothing on hand fixes it. Figma’s MCP server gives layout, not rules. A Storybook is built for people who click. A prompt resets every session.

1shared colour, of 17 and 37
No system, run 1, 17 colours
No system, run 2, 37 colours

Open a component, set variant, tone and size with live controls, and use it as you see it. The prop names in the table are the ones you type, and the ones an agent reads.

This is the real Button page, not a recording. Change a control and the button above it changes.

AGENTS.md loads before the first line of code. It tells an agent which tokens to use, what’s off limits, and what to do when something is missing.

AGENTS.md
## The three layer ruleComponents touch semantic and component tokens only.## Negative constraintsNo raw hex. No primitives. No arbitrary margins.## Fallback protocolPattern missing? Compose it, then log the gap.

Cursor, Copilot and Codex read AGENTS.md directly. Claude Code reads it through a CLAUDE.md that points to it.

Flip the theme. Nothing here is a screenshot.

Contrast lives in the tokens. Teal-700 carries white text at 6.27:1. Teal-600 is 4.21:1 and fails, so the rule is encoded, not remembered.

Three token layers: raw values, values named by job, values for one component. Components never touch a raw value, so dark mode re-points one layer and everything follows.

Every decision was made in Figma first. Then I rebuilt the file so nothing was implied, and encoded it here.

Built on MUI, not themed from it.

MUI handles focus, keyboard and menus. Weave owns the tokens, the rules and the docs.

Scoped the gate instead of making it strict.

The strict version failed 173 stories, because a lone component has no main landmark. A gate that fails everything gets switched off.

Held the input border, for now.

Every input border sits at 1.75:1. WCAG asks for 3:1. The fix is one token, but it changes every form at once, so I’m checking each one first.

CURRENT, 1.75:1

Email address

WCAG MINIMUM, 3:1

Email address

Two runs, no system

Two runs, with Weave

Colours
Two runs, no system
1shared, of 17 and 37
Two runs, with Weave
21shared, of 23 and 26
Font sizes
Two runs, no system
4shared, of 5 and 11
Two runs, with Weave
7shared, of 7 and 7
Spacing values
Two runs, no system
6shared, of 6 and 11
Two runs, with Weave
10shared, of 10 and 12
Corner radii
Two runs, no system
3shared, of 4 and 10
Two runs, with Weave
4shared, of 4 and 4
Typeface
Two runs, no systemTwo different ones
Two runs, with WeaveThe same one

One dot per distinct value across the pair. Filled means both runs used it. Measured in the browser with the same script on all four screens. Spacing counts padding and gap.

NO SYSTEM, RUN 1

No system, run 1: an indigo dashboard named Chalkline

NO SYSTEM, RUN 2

No system, run 2: a different indigo dashboard, also named Chalkline

WEAVE, RUN 1

Weave, run 1: a teal dashboard built from Weave components

WEAVE, RUN 2

Weave, run 2: a second teal dashboard built from the same Weave components

Four runs, one prompt. Layout varies in both pairs because the brief was open. In the Weave pair, almost nothing else does.

“Both controls meet the 44px hit area even though they are drawn smaller.”

The controls were drawn at 32px, with nothing in the CSS to make the tap area bigger. A linter can’t catch that. There’s no code to flag. A design review can’t either. The docs say it’s handled.

The controls were drawn at 32px, with nothing in the CSS to make the tap area bigger. A linter can’t catch that. There’s no code to flag. A design review can’t either. The docs say it’s handled.

Try tapping both. On a phone you’ll miss one.
32PX TAP AREA
Tap the button0 hits, 0 misses
44PX TAP AREA
Tap the button0 hits, 0 misses
This is the bug, not a diagram of it.

41

41

FAILURES REPORTED

39

39

IMAGINARY

The auditor made the same mistake twice. It checked see-through dark mode colours against a white background and reported contrast failures that didn’t exist. It caught the error, logged it and explained the cause. Next audit, same mistake.

Knowing about a mistake doesn’t stop you repeating it. Turns out that applies to machines about as well as it applies to me.

Auditing finds what contradicts itself. Building finds what’s missing.

Auditing finds what contradicts itself. Building finds what’s missing.

Agents built a working dashboard with nothing but Weave, and I logged every place it fell short.

18GAPS FOUND
10FIXED
8OPEN

Hardcoded widths. Card was fixed at 600px and ListItem at 260px. Both now fill their parent.

Missing pieces. No calendar and no stat tile. Both builds made their own, and they came out different.

Of the 8 still open, six are docs that describe a slightly different system.

01

Drift hides in the docs.

I documented a rule as met and never built it. No review process I’ve worked in would have caught that.

02

Scoping beats strictness.

A gate that blocks everything gets switched off.

03

Automated findings need checking, not trust.

I read every finding against the source before accepting it.

Next: close the 8 open gaps, decide the input border, and hand AGENTS.md to people who didn’t write it.