Skip to main content
inkbridge

Documentation

The methodology, the setup, and every part of the loop — broken down per topic.

Get started in 3 steps

Full setup guide →

From zero to a compiled design system — rendered in Figma and readable by AI agents — one install command, no manual exports.

  1. 1

    Install the npm package

    In your project root run pnpm add -D inkbridge then pnpm exec inkbridge setup. Setup creates the scanner + token API endpoints and gitignores the scanner output. No package.json scripts are added — your normal pnpm dev serves them.

  2. 2

    Load the plugin in Figma Desktop

    Go to Plugins → Development → Import plugin from manifest and select node_modules/inkbridge/manifest.json from your project. Figma remembers it — you only do this once.

  3. 3

    Generate your design system

    Run pnpm dev to start your dev server, then click Generate Design System Page in the plugin — colour tokens, typography, and every component variant are built on demand.

Methodology

How the story sits between design and code as the canonical artifact — and how each part of the loop works in practice.

Getting Started

Install Inkbridge and wire it into your Next.js project.

Component Scanning

How Inkbridge discovers your components and Storybook stories from inkbridge.config.json.

Design System

Extract tokens, themes, and styles from your Tailwind config — and push token edits back as PRs.

Figma Output

Control how components, previews, and states are rendered in Figma.

Agents / MCP

Expose your design-system model to AI agents over MCP — read components, variants, and tokens with no Figma in the loop.

Pro Features

The active research track — pushing structural component edits from Figma back to your stories as code-edit PRs.