https://www.ericaknauss.com/blog daily 0.75 2026-01-19 https://www.ericaknauss.com/blog/2026/1/20/design-systems-what-they-are-and-why-you-need-one monthly 0.5 2026-01-19 https://www.ericaknauss.com/blog/2026/1/18/crafting-better-prompts-for-figma-make monthly 0.5 2026-01-19 https://images.squarespace-cdn.com/content/v1/5921af79e6f2e10af2918c3c/2acdadc4-12f5-49ed-a791-8932c38e37a7/screencapture-figma-make-OYd4x743LPPNu17dI29rQd-Review-Screen-Design-2026-01-18-23_30_36.png Blog - Crafting Prompts for Figma Make - This is what the initial prompt produced and this is exactly what I want at this stage. The layout prioritizes structure over styling. Content is clearly separated from actions. Status is visible without competing for attention. Feedback is grouped in a way that makes it obvious who said what and why it matters. Most importantly, the screen makes roles legible: as a primary reviewer, I can immediately see what I’m responsible for and what decisions I’m allowed to make. The hierarchy is doing the work. You’ll notice that actions are contained and intentional. Approval and request-changes live together, but they don’t overpower the content. Supporting details—review metadata, activity history, attachments—are present, but secondary. They’re accessible without pulling focus away from the task at hand. This is the point of a good first pass with Figma Make. It’s not about shipping UI. It’s about validating that the shape of the solution makes sense before worrying about edge cases, refinements, or visual nuance. Interactive example https://images.squarespace-cdn.com/content/v1/5921af79e6f2e10af2918c3c/3bd83b8c-66ce-4da0-b209-6e4d68278b12/screencapture-figma-make-QynBFvEusmNp60MP35aUUH-Review-Screen-Design-2026-01-18-23_39_21.png Blog - Crafting Prompts for Figma Make - This is what happens when the prompt stays high-level and outcome-oriented. This is what happens when the prompt stays high-level and outcome-oriented. At first glance, the screen looks fine. Content is visible. Comments exist. Actions are present. But the moment you try to use it, the cracks show. Roles aren’t clearly differentiated, so actions feel generic rather than intentional. Approval controls appear without enough context, and it’s not immediately clear who is allowed to do what—or why. The hierarchy reflects the prompt’s ambiguity. Because the instructions didn’t define states, permissions, or decision boundaries, the interface defaults to a one-size-fits-all layout. Everything is technically there, but nothing is prioritized. This is the cost of an underspecified prompt. Figma Make filled in the gaps, but it did so by making assumptions you didn’t explicitly choose. The result is a UI that looks complete but is harder to reason about, harder to critique, and harder to evolve. https://images.squarespace-cdn.com/content/v1/5921af79e6f2e10af2918c3c/5baa4f15-5a8f-40a5-954e-de617a7a0219/screencapture-figma-make-58MTJgZpa9DRFyIKp0e0j8-Review-Screen-Design-Review-State-2026-01-18-23_43_25.png Blog - Crafting Prompts for Figma Make - This is the same screen (from the 1st initial prompt), but now it understands state. By asking Figma Make to generate variants for pending, changes requested, approved, and view-only states, the UI shifts from a static layout to a system that reflects how review actually works over time. Actions are no longer just present—they’re contextual. What I can do depends on both my role and where the item is in the process. You can see how hierarchy adjusts subtly across states. Primary actions appear when a decision is required and recede when it’s not. Status indicators do real work instead of acting as labels. In view-only mode, the interface de-emphasizes decisions entirely and focuses on comprehension. This is the point where the design stops being a screen and starts behaving like a workflow. Nothing new was added visually. The difference comes from making state explicit in the prompt. If the earlier version answered what is this screen, this version answers when am I here and what can I do now. Interactive example https://images.squarespace-cdn.com/content/v1/5921af79e6f2e10af2918c3c/6c5966ec-7b62-4ea4-b8a2-3bc8a5ae5f6e/screencapture-figma-make-NabIRfhHcVFFnF5v9Uzyv5-Review-Screen-Design-Review-State-Role-States-2026-01-18-23_46_59.png Blog - Crafting Prompts for Figma Make - At this point, the screen understands not just state, but who the user is. By rendering role-specific variations, approver, commenter, and viewer, the interface stops treating permissions as an invisible rule and starts making them legible in the UI. Actions appear, move, or disappear entirely based on what the user is allowed to do. Commenting remains accessible where it makes sense. Decision-making tools surface only when responsibility exists. What’s important here is what didn’t change. The layout stays familiar. The content remains stable. What shifts is emphasis. For an approver, decisions are foregrounded. For a commenter, feedback becomes the primary interaction. For a viewer, the screen prioritizes understanding over action. This is where prompts move beyond layout and into behavior. The design isn’t just reacting to data anymore, it’s responding to intent. Interactive example https://www.ericaknauss.com/blog/2026/1/17/ai-design-how-i-approach-figma-make monthly 0.5 2026-01-17 https://www.ericaknauss.com/home daily 1.0 2026-01-28 https://images.squarespace-cdn.com/content/v1/5921af79e6f2e10af2918c3c/f11a14e9-1917-4c6d-8173-dc1bd32b2060/ehk2.png