Ben Blumenrose pushes back on what the whole thread is really arguing about: ​“Product designers do help with UI design systems but that’s a fraction of what they do which also includes what product to build, how it should work, how to ensure people understand how to use it, how to create a brand people love and rally around, how the system should work together, etc etc etc but yes they’ll need to do visual design of UI design systems less so I’ll concede on that.”

“The first casualty won’t be design. It’s the deliverable.” Tommy Geoco backs it with numbers: a recent State of Prototyping survey of 1,478 design engineers finds 80.9% spending most of their week vibe coding, and 59% having shipped their own internal tool in the last six months. ​“The design function isn’t being eliminated. It’s absorbing engineering.”

Carina B. Velasquez, a Design Systems Lead locked out of AI tools at work, spent a Friday afternoon on her personal laptop with Claude Code and the Figma MCP server. By 6pm she had built Prisme: a complete three-tier token system, four brand themes, and a bidirectional plugin (Prisme Bridge) translating between Token Studio and native Figma variable formats. Her conclusion: ​“AI is amazing, Claude Code built my design system for me, the future is here. That is not what happened. What happened is that I showed up with years of accumulated knowledge, a specific problem I understood deeply, and a clear enough mental model to describe what I wanted with precision. Claude Code was fast; impressively fast, and smart in ways that genuinely surprised me. But it was executing a vision I already had. The intelligence in that Friday afternoon was not artificial.”

Dylan Field in an interview with Peter Yang argues that AI will massively expand what’s possible in design, but the enduring edge for humans is judgment: taste, craft, and point of view, plus the ability to use AI as an iterative partner rather than a one-shot generator.

Explore the world of Figma Weave workflows in the new Community area. Figma Weave (formerly Weavy) combines multiple AI models with professional editing tools in a node-based canvas, covering image, video, animation, and VFX.

Brad Frost and Chromatic cofounder Dominic Nguyen walk through how Storybook’s new MCP exposes design systems for agents to use reliably. The key insight: AI doesn’t invent states your components don’t have — it amplifies what’s documented. Coverage, completeness, and constraints are now load-bearing.

Steve Schoger from Tailwind Labs recorded a step-by-step walkthrough on taking an AI-designed dashboard and turning it into something that looks like it was designed by a pro.

Christine Vallaure built a five-person AI editorial team from one text file and seven lines. Five named agents: Rachel brings reader feedback, Joan brings numbers, Caitlin flags voice problems, Miranda raises buyer concerns, and Elke runs the room as editor in chief. Technically one model shifting between personas inside a single Claude conversation, but behaviorally a team critique. The most concrete personal account I’ve read of what agentic AI actually feels like to operate as a solo creator.

A thorough, practical guide from Christine Vallaure on what design systems need to work well with agents. Five things most systems are missing: semantic tokens, full state coverage, slot-based composition, exact Figma-to-code property matching, and token-based auto layout. Pairs well with the Brad Frost video — both cover the same design system moment from different angles.

Ed Bayes leads design on Codex at OpenAI; Gui Seiz leads AI product design at Figma. Together with Aakash Gupta, they give a rare behind-the-scenes look at what the Codex-to-Figma MCP integration actually changed inside their teams. The workflow is genuinely bidirectional, code and canvas reading from each other directly. The most concrete signal that something shifted: designers are working directly in staging.

Figma and Anthropic demonstrate the Claude Code roundtrip in both directions. Write to Canvas lets Claude Code create and modify designs directly in Figma; Code to Canvas brings live browser UI back in as editable frames. Worth watching alongside the Agents on Canvas blog post and Figma Skills page for the full picture of what this workflow ecosystem looks like.

Figma Skills are Markdown instruction files that teach AI agents how to reliably handle specific Figma tasks — generating a design system from a codebase, building screens from existing components, and more. Community contributors are already adding new ones.

Figma opens the canvas to agents. The use_figma MCP tool lets Claude Code and Codex generate and modify designs grounded in your actual design system. The key distinction from earlier code-to-design experiments: agents work with what your team has already built, making design system quality a direct input to AI output quality. 

Yann-Edern Gillet, design engineer at Linear, revisits his ​“Rosetta Stone” metaphor for design engineering translation through the lens of AI. The central argument: when translation is cheap and instantaneous, the bottleneck shifts from execution to meaning — and the new craft is preserving intent while everything accelerates.

Pair the Emil Kowalski skill with the Jakub Krehel skill above, and you’ve got a solid baseline of design engineering taste baked into your AI coding workflow. Emil’s focus is specifically on animations — spring physics, performance, and interruptibility.

Admins can now purchase additional AI credits in Figma, either as a subscription with volume-based pricing or pay-as-you-go for variable usage. Available now for Organization and Enterprise plans, coming to Professional in May.

Ridd interviews Kris Puckett, design manager at Stripe and formerly design lead at Mercury and Dropbox, on becoming an AI-native designer by using tools like Claude Code, OpenClaw, and personal AI to ship his own apps, supercharge his design practice, and build highly personal systems for life and work. I’m deeply curious about the last part and can’t wait to dig into his course Neuma on building a personal AI system.

The new Workflow Lab format, showing an end-to-end process, is a smart way to frame the new AI image tools in context. The three new tools (erase object, isolate object, expand image) are genuinely useful for anyone who’s had to leave Figma to do basic cleanup in Photoshop, and Vectorize finally removes a step that’s been a quiet annoyance for years.

Patrick Morgan makes a clean distinction that vibe-coding discourse keeps blurring: prototype code is for exploration, production code is for endurance. He is building a protected prototyping environment using Claude Code, a place where his team can move fast and then deliberately port the right assets across the boundary into production.

There is a clear parallel with how the design team at Notion works. In the recent episode of How I AI, Brian Lovin showed their collaborative ​“prototype playground,” where the entire team can create, share, and iterate on functional prototypes.

That also reminded me of how my team worked a decade ago, back when front-end development was a tad simpler. We had a separate ​“mockups” directory inside the Rails monorepo, where designers prepared static HTML mockups with production-ready CSS and JS. By the time designs were handed off to engineers in a feature branch, all polish and design details were already baked in. The design team must be fairly technical, but there is no going back to handing off Figma files after working this way.

Martin Rariga, Founding Product Designer at Equals, breaks down a practical AI-powered design engineering stack: Claude Code for big refactors and complex builds, Cursor for precise edits, and Figma (via MCP) to reference the new design direction while restyling.