Daniel Howells

Introducing Layout Mode for Agentation, a new way to explore and wireframe directly on the page. Rearrange and resize existing elements, add new components, and generate structured design feedback for your agent.

Over the past two months, my most impactful move as a design manager has been encouraging and supporting every designer to start building with Claude Code. Designers used to create a “source of truth” in Figma which was used for QA and accountability.. if anything was wrong, you’d point to the designs and say, “This is how it’s supposed to look.” The designers would take this artifact hand it off to engineers to build another “source of truth” on GitHub that would become the repo where other engineers can fork and build on top of. Now — the designer creates the source of truth on GitHub, and it’s closer to “the designs” because it is the designs. Less gets lost in translation, and everyone speaks in code. Figma remains useful for napkin sketches and quick visual experiments, but we are clearly moving beyond Figma as the primary document. The hardest part is getting set up—it’s daunting and scary leaving the things you love behind. People just need a little bit of emotional support to get started. But after taking the leap, everyone is feeling empowered and excited. They are all literally 10x’ing their productivity. The worst part is that engineers are a bit overwhelmed because they have way more PR’s to review.

Interesting. Gamma paid “the design tax” to a fantastic design agency for their brand identity. I’d love to see a robust public conversation about how AI design tools aren’t good enough for companies to use on their own brands, but are ostensibly good enough for their customers.

Anthropic is working on Sketch tool for Claude (as attachment)

Sorry kids, daddy is so upset by a github fork he’s spending the afternoon writing a Twitter blog about it. I used to really like Vercel. But why must their ceo and cto be so damn sensitive.

The token cost to build a production feature is now lower than the meeting cost to discuss building that feature. Let me rephrase. It is literally cheaper to build the thing and see if it works than to have a 30 minute planning meeting about whether you should build it. It’s wild when you think about it. This completely inverts how you should run a software organization. The planning layer becomes the bottleneck because the building layer is essentially free. The cost of code has dropped to essentially 0. The rational response is to eliminate planning for anything that can be tested empirically. Don’t debate whether a feature will work. Just build it in 2 hours, measure it with a group of customers, and then decide to kill or keep it. I saw a startup operating this way and their build velocity is up 20x. Decision quality is up because every decision is informed by a real prototype, not a slide deck and an expensive meeting. We went from “move fast and break things” to “move fast and build everything.” The planning industrial complex is dead. Thank god.