A breakdown of the new Claude Code to Figma workflow and what it means for design, code, and iteration speed at Prism.
share this post
stay in the loop
When we publish new experiments or playbooks, we’ll send you the highlights so you can apply them faster.
Your feedback helps us improve how we deliver practical playbooks.
Productized execution
Execution speed matters when your team is asked to ship often. Prism combines AI-assisted production workflows with stable web engineering and QA systems.
Keep learning
More experiments and playbooks from the Prism team.
from strategy to shipped ui: fast pages, clear messaging, and measured outcomes.
Today feels like a turning point.
Figma (opens in a new tab) just announced Claude Code (opens in a new tab) to Figma via MCP (opens in a new tab). You can now generate work in Claude Code, type "Send this to Figma," and the browser rendered state translates into fully editable Figma layers.
That sounds simple.
It is not.
It is architectural.
For years, product development followed a predictable path.
Idea to Design (opens in a new tab) to Code to Ship
Then AI arrived.
Suddenly we could:
Prompt to Generate to Iterate to Deploy
The speed was intoxicating. I have been living inside Claude Code and Codex for months, building systems, shipping experiments, and deploying apps in days instead of weeks. The acceleration is real.
But something subtle happened.
I drifted away from Figma.
Not because it was not powerful. Not because I stopped valuing design. The center of gravity shifted to the terminal. When you can generate UI instantly, you start designing in code by default.
The problem:
Momentum can become tunnel vision.
The first version becomes the version.
I have been using Figma since college.
Back when it felt revolutionary just to collaborate in real time. Back when Auto Layout felt like magic. Figma shaped how I think about product.
At its core, Figma is not about pixels. It is about perspective.
A canvas lets you:
AI expands possibility. Taste filters possibility.
In a world where models can generate nearly anything, the differentiator is no longer production capacity.
It is discernment.
The last two years were about intelligence.
Bigger models. Better reasoning. Faster inference.
Now we are entering the agency era. Tools that act, execute, and build.
Claude Code and Codex gave us execution speed. We can scaffold products instantly, wire up APIs, and deploy infrastructure.
This new Figma bridge adds something critical:
Reflection.
Instead of:
Prompt to Ship
We get:
Prompt to Generate to Send to Canvas to Explore to Refine to Pull Back to Code to Ship
That loop is powerful.
It reintroduces deliberate design into high speed building.
At Prism, we build growth (opens in a new tab) systems, AI tools, and high performance websites for ambitious teams.
Speed matters.
Quality matters too.
Over the past year, our iteration speed exploded. AI made it possible to test ideas quickly, deploy microsites, generate marketing flows, and spin up product experiments.
This integration changes the workflow again.
Here is how we will use it.
Instead of locking into the first generated layout, we will send early concepts to Figma and intentionally explore variations.
Not one landing page direction.
Five.
Side by side.
Then we choose.
When possibilities are infinite, clarity matters more.
The canvas becomes where we:
Previously, handoff felt directional.
Now it is cyclical.
Generate in Claude. Refine in Figma. Push changes back into code.
That loop compounds quality.
I love Claude Code. I love Codex. I genuinely believe they are the most powerful creative tools we have ever had.
I also love design.
Over the past six to twelve months, I felt myself drifting away from the design canvas because velocity was so addictive.
This feels like the moment that pulls everything together.
The terminal gives us leverage.
The canvas gives us judgment.
Together, they give us acceleration without sacrificing craft.
We are entering an era where:
The best builders will not be the ones who generate the most.
They will be the ones who navigate possibility space with precision.
That is why this matters.
Figma is not trying to compete with AI. It is positioning itself as the convergence layer: the place where ideas, prompts, sketches, terminals, and prototypes come together.
At Prism, we are doubling down on:
Building is accelerating.
Quality still matters.
The world we live in is unbelievable.
This integration feels like one of those subtle but important inflection points that, in hindsight, we will realize changed how software was made.
We are just getting started.
Enzo
