agentleFS
orbyn-devorbyn-dev/orbyn-fullstack-skills/skills/orbyn-ui-design/SKILL.md

orbyn-ui-design

Skill4 starsChanged 10 days agoMITView on GitHub

Design, implement, or review product UI and UX across web, desktop, and mobile: task flows, concise copy, responsive layouts, overlays, and AI interactions. Use for visible product flows; not backend-only work.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 12 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the agreed palette and shared components.

    line 9
  2. Distinctive:

    Read the consuming project's instructions and the agreed task scope.

    line 23
  3. Distinctive:

    Read design rules for typography, layout, copy, AI behavior, and the visual acceptance matrix.

    line 28
  4. Distinctive:

    Read UX rules when changing navigation, task steps, forms, onboarding, feedback, recovery, or explanatory content.

    line 30
  5. Distinctive:

    Read research when choosing or revising a rule.

    line 32
  6. Distinctive:

    Read the consuming checkout's design/brand pointers in its optional project context, then inspect the actual token and component owners.

    line 42
  7. Distinctive:

    Use existing text roles, spacing, and controls.

    line 55
  8. Distinctive:

    Follow the consuming project's visual-review policy and the changed surface.

    line 72
  9. Distinctive:

    Check the cases relevant to the named risk, then stop.

    line 76
  10. Distinctive:

    If browser permission or export is rejected, report the blocked action and use an allowed handoff.

    line 80
  11. Distinctive:

    When presenting new or changed UI for human design review, offer or reuse the live feedback method in review packages.

    line 83
  12. Distinctive:

    Read-only audit of an existing UI with plans for another agent: evidence-backed audit.

    line 97

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Orbyn UI and UX Design

Make the target product's task flow easy to scan and operate at real window sizes. Use the agreed palette and shared components. Compact default text must remain readable and grow with user accessibility settings.

This skill owns changes to the product UI itself. For social announcement graphics, product screenshot compositions and feed mockups, use social-preview, with social-content for the post; a promotional mockup is not an implemented screen. To measure motion against a reference and prove a build matches it, use motion-match. Copy or clone only our own work, or work the user owns or has licensed; for anyone else's product, measure how it moves and keep our own design. For faux product UI as a marketing still or cursor demo, use product-illustration.

Read before designing

  • Read the consuming project's instructions and the agreed task scope; preserve scope and stage order.
  • For product copy and information hierarchy, read orbyn-content-design; keep detailed content rules there.
  • Read design rules for typography, layout, copy, AI behavior, and the visual acceptance matrix.
  • Read UX rules when changing navigation, task steps, forms, onboarding, feedback, recovery, or explanatory content.
  • Read research when choosing or revising a rule. It distinguishes accessibility requirements, research guidance, and Orbyn choices.
  • Inspect the owning screen and shared controls before adding a local override. Resolve the theme/token and component owners from the consuming project's stack guide and manifests, and use the paths it already has. Never create paths to fit this skill. The Orbyn example shows one such mapping.

Resolve product direction

Read the consuming checkout's design/brand pointers in its optional project context, then inspect the actual token and component owners. Preserve its agreed palette, type scale, density, navigation and motion policy, and keep existing product tokens over a foreign host's theme or tool APIs. This library supplies no universal blue accent, modal layout or pixel scale. When a guide is absent, use the existing interface and the user's brief; record unsettled choices before a broad redesign. Audit-only requests produce findings without product edits. Keep a small polish to its stated scope.

Design and implement

  1. State the person's main task, starting state, and observable success. Map the steps, cancellation, failure, and return path before arranging controls.
  2. Use existing text roles, spacing, and controls. Reduce competing headings, repeated actions, and explanatory copy before changing font sizes.
  3. Adapt to the actual content width: collapse secondary panels, stack forms, or use a drawer. Keep every action reachable without squeezing the main task.
  4. Preserve data and authority: drafts, selections, permissions, approvals, consent, and backend run states must survive visual changes.
  5. Cover the supported surfaces affected by the requested change, using their platform-appropriate navigation and controls. Apply parity only where the product contract requires it; record agreed platform limitations.

Evaluate the complete task from the person's starting state through success, cancellation and recovery. Resolve trade-offs around that outcome, preserving accessibility and authority. Use the experience method when a local visual improvement could make the end-to-end task harder.

Review and deliver

Follow the consuming project's visual-review policy and the changed surface. Read visual review before capturing UI, briefing an inspector or handing a preview to a person.

  • Check the cases relevant to the named risk, then stop. Review density, discoverability and task completion as well as containment.
  • Fix reproduced problems, request a targeted recheck, and retain failed attempts and unverified states.
  • If browser permission or export is rejected, report the blocked action and use an allowed handoff. Never bypass it through another browser, port, transport or indirect export.
  • When presenting new or changed UI for human design review, offer or reuse the live feedback method in review packages.
  • For a requested prototype or preview, deliver the actual app route/story or a self-contained HTML artifact, with a verified link and the states to try. Small visual work stays small.

Read when the task needs it

  • Interactive sketch to resolve a design question: prototyping.
  • Unknown platform or accessibility requirements: orbyn-research.
  • Page titles, descriptions, canonical, social cards, icons or structured data: page metadata.
  • Documenting a product's design language for agents: DESIGN.md.
  • Read-only audit of an existing UI with plans for another agent: evidence-backed audit.
  • New visual direction, a requested redesign or a design-quality audit: visual direction.
  • Changing an existing UI: redesign contract.
  • Substantial workspace composition, collections, editors or review surfaces: workspace composition.
  • New application with no design system of its own: start from the Orbyn house style.
  • Colour direction, semantic light/dark roles or contrast review: reference palettes and semantic colour.
  • Validation with people: user evidence.
  • Model-powered behavior: AI application engineering and AI experience.
  • Assistant conversation/composer surfaces: assistant-chat UI.
  • Fetch/save/search or subscription changes: asynchronous state and recovery.
  • Stateful flows: behavior preview. Hand lifecycle statecharts to architect and critical property checks or drift to tdd.
  • Ordinary transitions/gestures: motion and measured motion performance.
  • Code-driven UI motion and an editable preview: motion composition and inspector.
  • Browser touch/keyboard/viewport: mobile web.
  • Accessibility audit or fix of controls, forms and dialogs: accessibility.
  • Real schema/data extremes: stress fixtures.
  • Demonstrated Expo/RN screens: native UI, navigation, data/lifecycle or UI-runtime motion, only as relevant.
  • Positive page identity, fresh structure after navigation, or supplied Figma node/assets-to-primitives mapping: design intake.
  • Bidirectional, translated or live-locale interfaces: localization and RTL.
  • A worked example: recovering a failed save applies a consuming project's style and draft/conflict contract without claiming an unperformed visual check.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.