shape

Overview of shape skill

What the shape skill does

shape is a planning-first skill for UI/UX Design. It helps you turn a rough feature idea into a design brief by running a structured discovery process before any code is written. The goal is not pixels or implementation details; it is to define the user problem, constraints, and product direction clearly enough that downstream work is faster and less guessy.

Who it is best for

Use the shape skill when you need a product-minded design starting point: new features, UX changes, redesigns, or any interface where requirements are still fuzzy. It is especially useful for teams that want an AI assistant to ask better questions before proposing layouts, flows, or component choices.

Why people install it

The main value of shape is decision quality. Instead of getting a generic UI concept, you get a brief shaped by context gathering, scope limits, and design intent. That makes it a better fit than a one-shot prompt when the real risk is building the wrong thing, not just building it slowly.

How to Use shape skill

shape install and setup

Install with npx skills add pbakaus/impeccable --skill shape. After install, open SKILL.md first, because the skill depends on the broader $impeccable guidance and asks you to follow its Context Gathering Protocol before making design calls. If your environment does not already include design context, run $impeccable teach first.

Start with the right input

The shape skill works best when you provide a feature goal, not a vague request like “make the UI better.” Strong inputs include the user goal, the product area, known constraints, and what success looks like. For example: “Shape the checkout flow for a mobile commerce app where guest users abandon at shipping selection.” That gives the skill something concrete to interrogate.

Suggested workflow for shape usage

Use shape in the planning phase: describe the feature, answer the discovery questions, then let the skill produce a design brief. Treat that brief as a handoff artifact for implementation skills such as $impeccable craft. If you already know the solution, shape is probably too early; if you do not know the problem well enough, it is exactly the right starting point.

Files to read first

Start with SKILL.md, then inspect the parts of the parent repo that define principles, context gathering, and handoff behavior. In this repo snapshot, SKILL.md is the only visible file, so the install decision hinges on its workflow rules rather than support scripts or examples. That means you should expect to supply more of the project context yourself.

shape skill FAQ

Is shape only for UI/UX Design?

Yes, shape is centered on UI/UX Design planning. It is not a code generation skill and it is not meant to produce final visual specs on its own. Its job is to clarify the design problem so implementation can follow with less rework.

How is shape different from a normal prompt?

A normal prompt usually jumps straight to a solution. The shape skill forces a discovery-first workflow, which is better when you need thoughtful product reasoning, not just a plausible mockup. If you want a prompt that immediately outputs screens, shape is not the best match.

Do beginners need prior design experience?

No, but beginners benefit most when they can describe the feature clearly. The skill can guide the conversation, yet it still depends on usable context: audience, goals, edge cases, and constraints. If you cannot answer those basics, expect the first pass to be incomplete.

When should I not use shape?

Do not use shape for tiny cosmetic tweaks, fully specified tasks, or late-stage implementation work. If the design direction is already locked and you only need execution, a build-focused skill is usually a better fit. Shape is strongest when uncertainty is the problem.

How to Improve shape skill

Give sharper context up front

The best way to improve shape results is to describe the user, the task, and the constraint surface in one paragraph. Include platform, audience, business goal, and anything that must not change. A weak prompt says “design a dashboard”; a stronger one says “design a B2B dashboard for ops managers who scan errors in under a minute, must work on laptop and tablet, and cannot add new nav items.”

Answer the discovery phase with specifics

The skill is built around context gathering, so vague answers degrade the brief quickly. When asked about goals or flows, answer with real behaviors, not preferences. For example, say what users do today, where they drop off, and what input data exists. That turns the shape skill from generic ideation into useful product direction.

Watch for common failure modes

The most common failure is skipping the interview and asking for a design too early. Another is giving output constraints without explaining why they exist, which leads to brittle recommendations. If the first brief feels broad, tighten the problem statement, clarify the primary user, and restate the success metric before iterating.

Iterate from brief to handoff

After the first output, refine the brief by adding edge cases, priority order, and any technical or brand constraints that surfaced. Then pass the improved brief to an implementation skill or your design workflow. That is where shape adds the most value: it reduces ambiguity before the work becomes expensive.