Maxime Rabot

Design System + AI

Independant visual designer & founder of House of Maestro

Set up the MCP bridge so Claude can read and update your tokens, components, and styles directly.

Connect Figma to Claude using the Figma MCP

Scaffold color scales, spacing, and type ramps — then catch naming conflicts and gaps automatically.

Generate and audit design tokens with AI

Prompt Claude to create and organize Figma components, then save those flows as reusable skills.

Build a Claude-powered component workflow

Du design au code avec Figma et Cursor. Exploitez le MCP Figma pour coder depuis vos écrans, user flows FigJam ou prototypes Make

Fondateur de Vibe Coding Academy

JULES BOITEUX

Fondateur de Vibe Coding Academy, plateforme d’e-learning sur le vibe coding

Fondateur AI Discipline

RÉMI GUYOT

Mars 26

Q2-2026

Les équipes Produit souffrent d'un goulot d'étranglement : trop de fonctionnalités restent dans Figma, faute de temps ou de ressources. L'IA change la donne en accélérant le passage du design au code. Dans ce cours, vous apprendrez à utiliser le serveur MCP Figma dans Cursor pour transformer vos écrans Figma en code prêt à être intégré. Vous verrez aussi comment générer des fonctionnalités à partir d'un user flow FigJam ou d'un prototype Figma Make. Nous utiliserons des écrans de démo, mais tout s'applique à vos designs existants, professionnels ou personnels. Que ce soit pour prototyper, amorcer un produit ou ajouter une fonctionnalité, vous découvrirez comment le serveur MCP Figma fluidifie le workflow design → code. Aucune compétence technique n'est requise. Vous maîtriserez Cursor et produirez un frontend professionnel sans être développeur. À la fin de ce cours, vous saurez : – Installer et configurer le serveur MCP Figma – Maîtriser les fonctionnalités principales de Cursor – Transformer un design (UI, FigJam, Make) en code fonctionnel – Itérer rapidement sur une fonctionnalité à partir de vos maquettes – Rebrander une application à partir d'un nouveau design system

Transformez un prototype Figma en fonctionnalité sans l'aide d’un développeur.

Intégrez un workflow design → code robuste et industrialisable à votre stack

Apprenez à appréhender les outils IA de Figma 

Apprenez à faire tourner une application en local

"Pas besoin d'une connaissance poussée de Figma, mais vous devriez avoir déjà appréhendé l'outil Figma 
pour designer des écrans"

Vous avez déjà créée des écrans dans Figma

Le serveur MCP de Figma est disponible sur tous les plans, y compris le gratuit. Le plan Pro est recommandé pour lever les limites d’usage.

Vous devez avoir créée un compte sur Figma

Vous devez avoir installé un IDE type Cursor, VS Code,... Version payante recommandée (~20€) pour utiliser les derniers modèles de LLM

Vous devez disposer d'un IDE type Cursor

Product DesignerProduct Manager / Product OwnerSoftware Engineer

✅ Livrez vos écrans Figma sans attendre✅ Transformer vos écrans Figma en code✅ Développer une app fidèle à votre DS

❌ Apprendre uniquement par la théorie❌ Vous refusez d'utiliser l'IA de Figma❌ Vous refusez d'utiliser Cursor

From Figma foundations to shipped product — in 5 weeks.

Founder moonlearning · 2× Config Speaker

Christine Vallaure

Founder, moonlearning.io · 2× Figma Config speaker · Awwwards Jury

Cohort 1

4-week hybrid course AI agents are reshaping how products get built. Designers who get this will have more creative control than ever. Designers who don't will find themselves further from the decisions that matter.Most courses on this are built for developers. This one is built especially for designers working in Figma, no engineering background needed.Weeks 1 to 3: Core Figma properly. Components, variables, auto layout, and naming. Understanding the foundation from the basics to the pro setup. Skip this, and all AI you add is built on sand and useless.Week 4: Get AI Agent ready: MCP and Code Connect demystified for designers, and a live exercise building your own lightweight agent with Claude Cowork.No fluff. No detours. Just the skills that matter right now.<hr> Includes:🧭 How it works, why it matters, design and code reality✍️ Hands-on exercises every week📌 Essentials first, extra mile optional🧩 One real project connecting everything🔀 Self-paced videos plus live sessions with Christine🚀 Week 4: Agentic AI for Designers, from MCP to making your own agent

The design skills that make you agent-ready

From pretty screens → a scalable UI foundation

From feature overwhelm → using Figma features with purpose

From handoff friction → dev & agent understanding

From Figma file →  Agent-ready AI design system

This course builds on basic Figma knowledge. You should know your way around the interface and core tools.

You should already be comfortable with:

Designers levelling up in FigmaYou know the basics but want real confidence in how your files are set up and why.

Designers who feel the AI shift comingUnderstand what agentic AI means for your work without wading through developer-focused content.

Designers working in teamsYou want a cleaner structure, better handoff, and a file that communicates intent clearly to humans and agents.

Go from trial-and-error prompting to systematic AI-assisted building. Ship functional projects with a workflow you can use on every project

AI Design Leader |  x-Uber/Google/Amazon

Stanford University

Amazon

Google

Uber

Rupa Chaturvedi

AI Design Leader | AI enablement for UX orgs | ex-Uber, Google, Amazon, Stanford

Cohort 2

The job posting asks for a designer who can research, prototype, code, and ship. One role, three people's worth of expectations. Design teams got smaller. The work did not. And now every product roadmap includes AI features that need to be designed by people who were never trained to design for probabilistic systems.The pressure is coming from two directions. Inside your organization, you are expected to move faster with fewer resources while somehow incorporating AI tools into workflows no one has figured out yet. Outside, the profession itself is shifting. AI generates interfaces, writes copy, produces prototypes in minutes. The execution work that built your career is compressing. The question is no longer whether designers need to adapt—it is whether you are building the skills that will matter in two years.Claude Code for Designers answers both pressures at once. You move faster because you can build functional prototypes yourself. You stay relevant because directing AI to build is the new execution skill. You leave with a deployed project and a workflow system you can use on anything that comes next.

Master Claude Code. Build functional projects. Develop a reusable workflow.

An AI-First Build Workflow: From project scoping through deployment

A Reusable Workflow System: Templates, skills, prompt libraries, MCP configs

A Deployed Application: Ship a functional project from your own idea

Portfolio artifacts and fluency to discuss AI-assisted design in interviews

Design-to-Code Translation: Turn your design system into functional components

This course assumes working knowledge of design systems, user research, and product workflows.

5+ years of professional design experience

Experience using AI assistants in your workflow. The course moves quickly and assumes baseline comfort.

Familiarity with AI tools (Claude, ChatGPT, Copilot)

Product Designers. You design in Figma and wait on engineering. You want to build and ship functional prototypes yourself.

Design Leaders. Your team needs to move faster and integrate AI workflows. You want firsthand fluency before rolling this out.

MXDAI Alumni. You have the strategic frameworks for AI design. Now you want the execution skill to match.

Learn to turn ideas into real products with Claude Code, generate PRDs, build MVPs, orchestrate agents, and ship deployable AI applications.

Founder | Ex-Google | Prof UCLA & UMN 

Walmart

Gallup

UCLA

University of Minnesota

Hamza Farooq

Founder | Ex-Google | Adjunct UCLA & UMN, SCU | Venture Partner

AI has changed how products get built. Tools like Claude Code are reshaping the product development life cycle in ways most of us are still catching up with.You probably have ideas. You see what these tools can do. But turning an idea into something real, a working product that runs, deploys, and can actually be used, can still feel overwhelming.The most exciting shift is that anyone today can go from idea to PRD to working prototype in minutes. That's a massive change in how products come to life.This course exists to help you take advantage of that shift.Instead of spending weeks watching videos, you'll build alongside us. You'll take an idea, turn it into a PRD, generate an MVP, orchestrate AI agents, and deploy real-world use case products.You won't just learn how to use Claude Code. You'll use it to build real products.🎯 Week 1: Turn idea to PRD into an MVP🎯 Week 2: Build your own marketing agency using sub-agents, and multi-agent orchestration🎯 Week 3: Evaluations 🎯 Week 4: Demo Day!By the end, you'll know how to go from idea to requirements to prototype to deployable product using AI-native workflows, and you'll leave with multiple working products you built yourself. 🚀

Turn ideas into deployed products using Claude Code, build 6 real applications 

Turn Ideas into Working Products

Build and Deploy Real Prototypes

Create Multi-Agent Workflows

Use MCP Connectors to Extend AI Apps

Build and Launch a YC-Style Product

Graduate with a Portfolio of Shipped Projects

Just be willing to learn, we will take care of the rest

Curiosity and openness

Product Managers &amp; Product Leaders who want to go beyond writing specs and start turning ideas into products using AI-native development.

Developers &amp; Builders who want to move faster with tools like Claude Code and build real prototypes, agents, and deployable AI applications.

Entrepreneurs &amp; Founders who want to validate ideas quickly, build MVPs, and launch solutions without waiting on long development cycles.