agentleFS
wicii2120wicii2120/skills/skills/vibe-design/SKILL.md

vibe-design

Skill1 starsChanged 3 months agoMITView on GitHub

Professional interface-design and design-system workflows. Use to clarify design terminology or intent; diagnose existing UI quality, consistency, usability, accessibility, content, or motion; systemize reusable foundations, tokens, components, governance, migration, or adoption; or craft concrete specifications, prototypes, visual directions, and production frontend implementations.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. aakashdhar/vibe-skill★ 7same name
    2. hexbee/hello-skills★ 3same name
    3. Cashmeran/hlvibes-skills★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository8 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    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 mode labels only when they clarify routing, sequencing, assumptions, or scope.

    line 17
  2. Distinctive:

    Ask only when an answer would change scope, visual direction, architecture, priority, fidelity, or acceptance criteria.

    line 24
  3. Distinctive:

    Gate: Every loaded file has a direct trigger in the current phase.

    line 69
  4. Distinctive:

    Use canonical English terms as semantic anchors.

    line 79
  5. Distinctive:

    Keep the core portable: semantic HTML/CSS concepts, platform role trees, typed token data, state models, contracts, pseudocode, and …

    line 81
  6. Distinctive:

    Treat normative accessibility as acceptance criteria and stricter thresholds as product requirements.

    line 83
  7. Distinctive:

    Keep evidence, assumptions, normative requirements, product constraints, governed rules, contextual heuristics, and exploration options …

    line 84
  8. Distinctive:

    Check every requested output and acceptance criterion against rendered or inspectable evidence where available.

    line 89

≈ “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

Vibe Design

Turn interface intent into grounded design decisions, portable contracts, and verified artifacts. Route by the object being changed:

Knowledge or intent            → Clarify
Existing artifact or behavior  → Diagnose
Reusable system capability     → Systemize
Bounded concrete artifact      → Craft

Use mode labels only when they clarify routing, sequencing, assumptions, or scope.

1. Frame

  1. Inspect supplied artifacts and the project before asking about them.
  2. Select one mode or an explicit chain. Default “improve” and “redesign” of existing UI to Diagnose → Craft unless the user supplies a trusted diagnosis or asks to skip assessment. Valid chains include Clarify → Craft, Diagnose → Craft, Diagnose → Systemize → Craft, and Systemize → Craft.
  3. Record outcome, product/platform context, audience, evidence, existing system/assets/content, and material brand, accessibility, performance, browser/device, technical, dependency, and tool constraints.
  4. Ask only when an answer would change scope, visual direction, architecture, priority, fidelity, or acceptance criteria. Otherwise state an assumption and proceed. Routine implementation needs no ceremonial approval.
  5. Route token architectures and component contracts to Systemize when ownership, governance, migration, adoption, or cross-product architecture is unsettled; route a bounded artifact from settled decisions to Craft; chain both when needed.

Gate: Each phase has an object, outcome, evidence boundary, constraints, assumptions, and a checkable completion condition.

2. Select references

Load exactly the current phase’s workflow. Add only references triggered by in-scope decisions. In Craft, every foundation explicitly named by the artifact fields is required even when its decisions are settled; settled means apply rather than reopen.

Mode triggerRead
Explain, define, compare, name, or translate intent into canonical termsClarify
Audit, assess, inspect, critique, or identify problems in existing evidenceDiagnose
Create or evolve reusable foundations, tokens, components, governance, migration, or adoptionSystemize
Design, prototype, specify, implement, redesign, polish, or emit a bounded artifactCraft
Cross-cutting triggerRead
Concrete visual direction, composition, aesthetic judgment, craft critique, Design Read, or anti-default reviewVisual Direction & Taste
Named style, style comparison or explanation, style-led exploration, or unresolved direction where named alternatives reduce decision riskVisual Style Lexicon
Direction is unresolved, alternatives reduce decision risk, or the user requests variantsPrototyping & Variant Exploration
Runnable UI, code change, rendering, screenshot inspection, or output-completeness obligationImplementation & Visual Iteration
A token file, contract, state matrix, specification, brief, theme, reference-image plan, motion review, or migration artifact is requestedDeliverable Templates
Foundation triggerRead
Typeface, text style, hierarchy, measure, type loading, or type personalityTypography
Palette, color role, contrast, mode, gamut, surface, or color tokenColor System
Composition, container, grid, responsive hierarchy, reflow, positioning, or overflowGrid & Layout
Gap, inset, rhythm, density, control size, or target separationSpacing
Photography, illustration, diagram, crop, generated asset, responsive source, or alternative textImagery
Icon meaning, family, SVG, label, badge, optical size, or hit areaIconography
Interaction state, feedback, animation, gesture, spring, interruption, or reduced motionMotion & Interaction
Object naming, labels, helper text, interface voice, state messages, claims, or localizationContent & UX Writing
Adapter triggerRead
Browser-native HTML/CSS/JavaScript or WAAPI is supplied or detectedBrowser-native Adapter
React, Next, Tailwind, Motion, or GSAP is supplied or detectedReact-stack Adapter
Native mobile, Expo, Reanimated, Gesture Handler, or haptics is supplied or detectedNative-mobile Adapter
Browser rendering, interaction inspection, screenshots, accessibility tree, or Playwright/equivalent is available or requestedBrowser Inspection Adapter
Image generation or semantic visual-artifact tooling is available and usefulVisual Tools Adapter

For citations, source conflicts, or licensing, read Provenance. A pinned-source refresh or exhaustive source-disposition request is a maintenance branch rather than a runtime design mode: read only Provenance and Upstream Coverage.

Gate: Every loaded file has a direct trigger in the current phase; intentionally unselected workflows and adapters remain unloaded.

3. Execute

Apply the selected workflow contract to its completion gate. In a chain, finish and verify one phase before passing its evidence and decisions to the next. Keep a local Craft request local; introduce Systemize only when reusable capability is part of the stated outcome.

Gate: The phase returns its required decision, finding register, system plan, or bounded artifact with acceptance evidence.

4. Apply invariants

  • Use canonical English terms as semantic anchors.
  • Resolve conflicts in this order: license and normative standards → explicit user/product/brand/technical constraints → mode contract → multi-source principles → source-specific craft heuristics → examples and preferences.
  • Keep the core portable: semantic HTML/CSS concepts, platform role trees, typed token data, state models, contracts, pseudocode, and behavioral checks. Adapt only to a supplied or detected stack.
  • Ground visual choices in audience, task, content, brand, and platform. Treat familiar patterns as evidence questions, not automatic defects; novelty without purpose has no value.
  • Treat normative accessibility as acceptance criteria and stricter thresholds as product requirements. Dark mode, a visual style, a design system, and a motion engine require explicit product evidence rather than default adoption.
  • Keep evidence, assumptions, normative requirements, product constraints, governed rules, contextual heuristics, and exploration options distinguishable.
  • Preserve complete vertical slices and an output ledger; requested work is implemented rather than represented by placeholders or silent omissions.

5. Verify and close

  1. Check every requested output and acceptance criterion against rendered or inspectable evidence where available.
  2. For runnable UI, complete the render/inspect/refine loop in Implementation & Visual Iteration.
  3. Report concrete blockers and the strongest available fallback for unavailable tools or untestable behavior.
  4. Return the checkable artifact or result, then prioritized Now / Next / Later actions with owners or dependencies when known.

Completion: The result is grounded, portable at its contract boundary, accessible, complete, traceable to evidence or explicit assumptions, and verified at the level the available artifacts and tools permit.

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.