New Tools for a New Era. Coding agents like Claude and Cursor have dramatically reduced the time it takes to go from idea to functional software. But the experience of designing and refining with them sucks. One reason for this is that while the terminal is an incredible tool... for communicating direction with language, it is a terrible tool for defining and exploring visual and interactive objects. Here is one idea for how we might fix a small part of that. In the old world, when you wanted to create a transition or animation in your app, you would type some code, refresh your local server, and click to run your animation. It probably wasn't right, because after all no one can know what 'cubic-bezier(0.3, 0.05, 0.45, 1)' really feels like when you read it. You need to see it. Feel it. Interact with it in a real world context. So you'd edit some values, save, refresh, and keep guessing and checking until it felt right. Today, you can write a quick, single-use tool that's a visual studio for designing animations. You can then configure some components and containers common in all apps, and explore different animations in real time, adjusting key properties, and getting it just right. Then, you can copy a highly detailed prompt (or export a skill containing all your animations) that captures your intent and direction with perfect clarity. Paste this into your terminal and your agent instantly implements it everywhere. To me, this is an improvement over the old world, and a better way to work in today's. I'm extremely excited to see the ways in which our ability to rapidly create software will shape how we design software tomorrow. Feedback, ideas, and critiques welcome!show more

Also, I think the “copy prompt” pattern is going to explode. Incredible DX.

got this—very roughly—working with Base UI this morning 🤔

Josh, my team and I have been building things JUST like this. The speed at which you can spin up a sandbox for exploring the way that any given code will work, look, and feel is unreal. Timelines are collapsing with the barriers to entry. Absolutely love how quickly we can all move now!

nice! excited to see what y'all ship here. P.S. Y'all/@zoink should sponsor Tailwind! Love that Figma's already sponsoring Motion.

@zoink To be clear, “my team” is referring to the NAMER Advocacy crew.

@zoink one team, one dream, right?

@zoink Indeed

Refining still has too much *cognitive* friction Human-in-the loop is the sweat spot for now ;)

This is the way I'm pretty bullish on the idea that we will build all of our stuff as part of the same flywheel. An operating system The UI and motion system, and the tools to maintain and adjust things are all baked in. Themeable, hackable, malleable That's at least what I'm investing all the time building 😂

Ya, I do question the longstanding decision of most headless libraries to stay agnostic (to some extent). That *might* have been appropriate in a time and a place, but every system I've designed internally at larger co's (Dropbox, Wealthfront, etc.) has been all encompassing. Seems things will trend that way as well

Yea it's difficult to say But I do love your approach above. I found myself trying to solve things visually without the "code and refresh" approach One thing I've been building through my operating system project is bi-directional visual and code editing with the ability to edit on top of production Although I love code, I want freeform exploration to refine. No one has ever done that in a way that made sense to me. AI will probably help us build these tools

Wil def do this for the design system I’m working on right now. So far started with building a custom variable explorer and editor where I can play around with my design variables which is already awesome.

Excited to see it

so good! are you planning on open sourcing this? would love to play around!

There’s been a lot of interest, so I might release it in some way.

Bullish on your take👌🏻 Idk wich software you are using but it looks really cool for iteration and fine tuning animations and small details.

thanks! It's homecooked software I made with claude. Working name: AnimationKit. Should i release it?

You should release it yesterday! And if you're not planning to charge, I would open source it immediately. It would go viral pretty quickly and give you great exposure.

What’s this app? Link?! 👀

one I'm just toying with right now. Should i release it?

Listen to the people - do iiiiit

Such a great idea! Once the prompt is copied over you can just ask the model to record to context or even better use it to establish a motion system in your project to have consistency across all interactions. So cool thanks for sharing!

Yeah, the ability to manually make edits with the mouse in conjunction with the ability to prompt in the same interface is a great combo IMHO. Sky is the limit. Great work 🤙

This is dope man!

cheers Sam!

Love this Josh!

Thanks Noah! Just trying to further conversations

Totally agree! The speed from idea to working code is incredible now. For the design/refinement UX gap, I've been combining Claude Code with component libraries and it helps a lot. The iteration cycle is still way faster than before even with some rough edges 🚀

i’d use it 100%

Ok but I just have to say your PFP took me like way too longer to figure out 🙈

ahah i promise, no ai involved! just Utah nature:)

Don't you think that terminal was made for developers not for "idea guys" ?

design is something that significantly lacks in ai and always questioned why that was the case because i am pretty sure it was trained on some great looking software, but the touch and feel part explains a lot.

Dope! Would love to play with it!

I have been building something similar and it takes this same direction

This is really cool! Please release it, hopefully open source 🙏. Would love to play around with it and bring in my own components at some point.

pleaes release it, can we try it?

This is brilliant UX! Would love to explore integration with @matterhornso

crazy

this mirrors how i've been building animation playgrounds for each project. the "guess and check" loop always felt broken. visual tools that export intent > trying to decode easing curves in your head

you get it 👌

Completely agree! I wish Figma MCP was good enough to create a visual and attach it to a flowchart data model with a description of what you want, and that’d be ingested by the AI in order to produce code.

This is amazing. And so useful for designers learning the ins and outs of coding with motion. Would love to be an early adopter of this!

Love the idea. Feels like there’s a lot of tools that needs to be cooked around design and AI! 👍🏻

Are you also looking at manipulating the GUI? Like many, I find the pixel pushing stage of vibe coding a drain.

I've written an article about this topic in last year. The human brain is naturally wired for visual and motor based interactions rather than structured linguistic articulation. Love your tool concept, hope you can release it to test.

This resonates hard. Built jsOM to solve the same pain: Visual canvas → LLM-readable spec → agent builds it. No screenshots. No guessing. Open source.

Wow 😮

I took a shot at building exactly this. Here is Tangent – a bi-directional visual tuner for AI-generated code. Open source here:

You should share the prompt, or throw this on GitHub pages and let some people beta it

Looks super interesting, release it so we can try it out and help you iterate — those of us backend folks who are forced to do front-end stuff will really appreciate it