agentleFS
cleissonbarbosacleissonbarbosa/ui-humanizer/SKILL.md

ui-humanizer

Skill1 starsChanged 8 months agoMITView on GitHub

Transform AI-generated web UI into authentic, professional interfaces with human character. Use when reviewing or refactoring frontend code that looks generic, template-driven, or "AI-generated". Triggers on: "humanize UI", "fix generic design", "make UI look professional", "remove AI look from interface", "add personality to UI", "de-templatize design", "authentic design". Applies the 5-Layer Authenticity Stack to detect and fix visual, interaction, layout, copy, and UX red flags in HTML/CSS/JS/React/Vue/Svelte code.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. gabrielgreco1/skills★ 98same 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.

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 16 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈16

    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:

    After generating UI with AI tools (Copilot, v0, Bolt, etc.)

    line 23
  2. Distinctive:

    When a design "feels generic" but you can't pinpoint why

    line 24
  3. Distinctive:

    When refactoring a prototype into production UI

    line 26
  4. Distinctive:

    When adding brand personality to a white-label template

    line 27
  5. Distinctive:

    Read all target files before making any changes.

    line 31
  6. Distinctive:

    Check whether a design system or token file already exists in the project.

    line 33
  7. Distinctive:

    Read the target files, then scan for these telltale signs of AI-generated UI:

    line 40
  8. Distinctive:

    Layer 2 — Micro-Interactions: Add purposeful motion that provides feedback and personality (hover lift, focus rings, skeleton loading, …

    line 61
  9. Distinctive:

    Start with the highest-severity flags from Phase 1.

    line 79
  10. Distinctive:

    Test after each layer — visually verify changes do not break layout.

    line 83
  11. Distinctive:

    Before marking complete, verify:

    line 89
  12. Distinctive:

    Never change functionality or business logic — only presentation and interaction.

    line 106
  13. Distinctive:

    If a design system already exists in the project, extend it instead of replacing it.

    line 108
  14. Distinctive:

    Always check accessibility: contrast, keyboard nav, screen reader labels.

    line 110
  15. Distinctive:

    Do not add external dependencies without confirming with the user.

    line 111
  16. Distinctive:

    After completing the humanization, provide a summary:

    line 115

≈ “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

UI Humanizer

Transform generic, template-driven interfaces into authentic, professional designs that feel crafted by humans. Based on the 5-Layer Authenticity Stack methodology.

When to Use

  • After generating UI with AI tools (Copilot, v0, Bolt, etc.)
  • When a design "feels generic" but you can't pinpoint why
  • During design review of frontend components
  • When refactoring a prototype into production UI
  • When adding brand personality to a white-label template

Before Starting

  1. Read all target files before making any changes.
  2. Identify the framework in use (React, Vue, Svelte, plain HTML/CSS, etc.).
  3. Check whether a design system or token file already exists in the project.
  4. Confirm the scope: full page, single component, or set of components.

Procedure

Phase 1: Detect AI Red Flags

Read the target files, then scan for these telltale signs of AI-generated UI:

CategoryWhat to look for
VisualOverly perfect gradients, trending color combos (indigo-purple, blue-cyan), identical cards, uniform spacing, uncurated icon sets, same border-radius everywhere, no visual hierarchy
InteractionZero micro-interactions, generic hover (opacity: 0.8), bare spinners, no transitions on state changes, empty states that say "No data"
Code smellsSame spacing value repeated (p-4 gap-4 m-4), no design tokens or semantic colors, inconsistent inline styles vs utilities, no responsive adjustments, no clear type scale

Report findings organized by severity: Critical (breaks professionalism), Major (feels generic), Minor (missed polish opportunity).

For the full checklist of red flags with examples, read references/red-flags.md.


Phase 2: Apply the 5-Layer Authenticity Stack

Work through each layer sequentially on the flagged components. Read the detailed reference for each layer on demand.

Layer 1 — Visual Identity: Replace generic palette with semantic color tokens, establish type scale, create spacing and border-radius hierarchies. → Read references/layer-1-visual-identity.md

Layer 2 — Micro-Interactions: Add purposeful motion that provides feedback and personality (hover lift, focus rings, skeleton loading, contextual loading messages). → Read references/layer-2-micro-interactions.md

Layer 3 — Layout Intelligence: Break out of uniform grids with bento layouts, progressive disclosure, content-driven breakpoints, and intentional whitespace. → Read references/layer-3-layout-intelligence.md

Layer 4 — Copy Personality: Rewrite robotic UI text so empty states, errors, loading, and onboarding sound like a person wrote them. → Read references/layer-4-copy-personality.md

Layer 5 — UX Architecture: Evaluate emotional context, workflow fit, accessibility coverage, and brand personality alignment. → Read references/layer-5-ux-architecture.md

For the full methodology, design tokens, and rationale, read references/authenticity-stack.md.


Phase 3: Implement Changes

  1. Start with the highest-severity flags from Phase 1.
  2. Edit one component at a time — do not refactor everything at once.
  3. Preserve existing functionality — only change presentation and interaction.
  4. Apply changes in this order: tokens/variables → typography → spacing → colors → interactions → copy.
  5. Test after each layer — visually verify changes do not break layout.

Phase 4: Polish Checklist

Before marking complete, verify:

  • Semantic color tokens instead of raw color values
  • Clear typography scale with visual hierarchy
  • Intentional whitespace with visible rhythm (no uniform spacing anti-pattern)
  • Consistent border-radius hierarchy across components
  • At least 3 distinct micro-interactions (hover, focus, loading/empty state)
  • Focus states visible and branded
  • Human-sounding copy on empty states, errors, and loading
  • Custom or curated icon set (not raw default library dump)
  • Accessibility: contrast ratios pass (4.5:1 text, 3:1 large text/UI), keyboard navigation works
  • prefers-reduced-motion respected where animations are added

Rules & Constraints

  • Never change functionality or business logic — only presentation and interaction.
  • Preserve all existing class names, IDs, and data attributes that other code may depend on.
  • If a design system already exists in the project, extend it instead of replacing it.
  • Respect the framework's idioms (e.g., use className in React, class: in Svelte).
  • Always check accessibility: contrast, keyboard nav, screen reader labels.
  • Do not add external dependencies without confirming with the user.

Output Format

After completing the humanization, provide a summary:

## UI Humanization Report

### Red Flags Found
- [severity] description → fix applied

### Layers Applied
1. Visual Identity: [changes made]
2. Micro-Interactions: [changes made]
3. Layout Intelligence: [changes made]
4. Copy Personality: [changes made]
5. UX Architecture: [changes made]

### Before/After
[key visual differences]

### Remaining Opportunities
[optional polish items for future iteration]

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.