frontend-design or canvas-design: which one fits the work?
frontend-design is for interfaces that will ship as code, with the taste applied to components, layout and states. canvas-design is for composed visual work on a canvas. If the output ends up in a repository, you want frontend-design.
The frontend-design skill gives Claude explicit design direction so generated interfaces stop converging on the same centered card, gradient header and rounded everything. It changes the starting point, not your taste. You still have to reject the first result when it is boring.
Best for
Landing pages, dashboards, prototypes and artifacts where the look is part of the point.
Skip if
You are working inside an existing design system, where the components already decide the visual language.
Which one to install
Choose frontend-design when the deliverable is a page or a component that someone will run, and the risk you are managing is that generated interfaces come out generic. It works at the level of the code that gets committed.
The canvas-design skill focuses on composition for visual pieces such as posters, covers, diagrams and graphics, where hierarchy and spacing carry the meaning. Use it when the artifact is an image rather than an interface, and reach for frontend-design when the artifact has buttons.
Best for
Posters, social graphics, covers, event artwork and any single composed image.
Skip if
You need an interactive interface, where interaction states matter more than composition.
Which one to install
Choose canvas-design when the deliverable is the image itself: a composition, a poster, a laid out surface where the arrangement is the product rather than a step towards the product.
They can be installed together without much interference, because one is triggered by building an interface and the other by composing a visual. The description to watch is any third design skill you add later that describes both.