A product is held together by a million little decisions, many of which have become muscle memory for designers. Debprotim Roy introduces the design harness, a way to make those decisions part of how an agent works.

Over a decade of product work, we have built design systems for project after project. Omni now sets up a working system for designers and agents in minutes, right from the get-go. Arjun Rajkishore walks through how he built it and how it works.

Canvs has built its practice inside complex products. We look at how agentic workflows put that accumulated understanding to work when the ask in front of us is to rethink an existing product or imagine a new one.

The most useful research question is often one that may not arise at the beginning of a project. Premankan Seal writes about building a persistent research system that follows questions as they emerge, draws on what earlier projects uncovered and brings evidence into decisions while they are still open.

A client privacy constraint nudged Himanshu into building his own tool for small file jobs. He talks about how he built it and how that maps to the way he builds with agents generally.

For more than a decade, Canvs has documented the choices and reasoning around its projects. We talk about the intelligent system we are building that is always learning and contextualizing what it picks up, bringing that history into decisions as they happen.

Building internal tools has become faster and cheaper than it has ever been, and we've seen this internally, and we see this outside. Debprotim writes about what's changed inside the studio in the last six months as Canvs itself went agentic.

Hamsika took an existing design system, already running across tens of files and hundreds of live flows, and explored what building a dark mode for it would take.