Small prototypes, prompts, and documentation experiments that test where AI and faster tooling help design judgement, and where they distort it.
How generative prototyping can turn early product questions into tangible sketches and interaction ideas.
The value is not automatic output. It is the shorter distance between a question and a useful artefact.
Patterns for how people and AI agents can share work, hand off context, and keep decisions understandable.
Agentic products need visible handoffs, not only conversational surfaces.
Documentation structures that help people and AI tools understand component intent, constraints, and usage.
A component needs purpose, behaviour, and boundaries, not only tokens and examples.
Using prompts to explore scenarios, edge cases, alternative flows, and product assumptions.
The best outputs come from clear product judgement before the prompt is written.
How design system knowledge can move between documentation, prototypes, and implementation prompts.
Translation is a design problem: the important part is preserving intent, not only appearance.