agentleFS
gordon1210gordon1210/agent-goodies/skills/godot-ui-design-excellence/SKILL.md

godot-ui-design-excellence

Skill0 starsChanged 4 months agoMITView on GitHub

Design, implement, critique, and polish distinctive game UI in Godot: visual art direction, HUDs, menus, inventories, strategy dashboards, typography, materials, motion, controller navigation, accessibility, and in-engine visual validation. Use for premium or AAA-quality interface requests, Expedition 33 / Frostpunk 2 references, UI redesigns, and interfaces that work but look generic. Not for website design, general gameplay implementation, or engine migration.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 7 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read only the relevant rows below.

    line 27
  2. Very specific:

    Build one representative screen: Use real or labeled fixture data, real background variation, and normal/focused/disabled/error states.

    line 71
  3. Distinctive:

    Implement natively: Use semantic Controls and reusable scenes, Themes and variants, authored assets, and bounded decorative effects.

    line 74
  4. Distinctive:

    Inspect and revise: Run the real scene, capture it at target sizes, exercise input and animation interruptions, then correct the most …

    line 77
  5. Distinctive:

    Use texture and motion for a stated purpose, not to conceal weak hierarchy.

    line 90
  6. Distinctive:

    Keep functional text and interaction native.

    line 91
  7. Distinctive:

    No blanket ban on manual coordinates, continuous updates, or linear motion.

    line 95
  8. Distinctive:

    No invented runtime tests, fabricated screenshots, invisible tool capabilities, or claims of accessibility certification.

    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

Godot UI Design Excellence

Produce a coherent, playable interface with an authored visual identity. "AAA" is an ambition for craft, not an aesthetic preset or a quality guarantee. Preserve the project's identity; never substitute a generic dashboard, fantasy skin, or unrequested redesign.

Start with the right amount of context

Inspect the current screen, project instructions, engine version, renderer, target devices, existing UI scenes/themes, and available tools. Read discovery for new work; skip repeated discovery for a small established task. Ask about genuinely unresolved design requirements rather than inventing the user's game. Do not ask questions already answered by the repo.

Read only the relevant rows below. Usually load 2–4 modules for the current stage, not the entire library. Source and example files are optional lookups. Finish a stage, persist decisions, then load the next stage's references. For a small fix, choose the narrow repair route without restarting art direction.

Task / symptomRead
New visual identity, references, "make it modern"Reference analysis, art direction
Generic layout, weak hierarchyComposition, critique and repair
Fonts, numbers, readabilityTypography, Godot text
Colors, texture, ornaments, iconsColor and materials, iconography
State language and reusable componentsComponent states, Godot themes
Information architecture and decisionsInteraction
Cinematic / painterly RPGRPG direction, worked RPG example
Dense management / strategyStrategy direction, worked strategy example
Expressive or tactical identityOther visual directions
Inventory, equipment, build comparisonInventory and buildcraft
Maps, research trees, spatial navigationMaps and trees
Tooltips, alerts, notificationsTooltips and notifications
Settings, confirmation, dialogueSettings and dialogue
Motion / transitions / soundMotion design, audio feedback
Art production / generated assetsAsset production
Layout, scaling, safe areasGodot layout
Input leak, focus, modal, device switchingGodot input
Interrupted animations, layout fighting tweensGodot animation
Shader, blur, masks, custom drawingGodot rendering
Portraits, 3D previews, world-space UIGodot spatial UI
Data binding, loading, large lists, performanceGodot data and performance
Accessibility or localization acceptanceAccessibility, Godot text
Verify or review a delivered screenVisual validation, critique and repair
Permissions, tools, external referencesEvidence and safety
Primary sources / changing APIsSource register

Production loop

  1. Frame the decision. Who is acting, what must they notice, what can they do, and what is the cost of a mistake? Capture constraints in the design brief, proportionate to the request.
  2. Observe references. Inspect actual screenshots and, for motion claims, recordings. Distinguish shipped screens, concepts, and original proposals. Extract a visual grammar, not a bag of decorative motifs.
  3. Choose a direction. Define composition, typography, shape/material roles, information hierarchy, and motion character. Explore meaningfully different options only when direction is open. Respect explicit delegation of choices.
  4. Build one representative screen. Use real or labeled fixture data, real background variation, and normal/focused/disabled/error states. Use the screen contract. Do not scale out an unproven skin.
  5. Implement natively. Use semantic Controls and reusable scenes, Themes and variants, authored assets, and bounded decorative effects. Detect the installed Godot version before using version-sensitive APIs; never upgrade it implicitly.
  6. Inspect and revise. Run the real scene, capture it at target sizes, exercise input and animation interruptions, then correct the most consequential visual or interaction failure. Use the visual review.
  7. Generalize after acceptance. Extract components and tokens from the accepted screen, apply to sibling screens, and record exceptions. Hand off evidence and remaining limits with the handoff.

Non-negotiable distinctions

  • Functional correctness, visual fidelity, usability, and accessibility are separate acceptance axes. A polished screenshot cannot compensate for a trapped user; a passing build cannot prove visual quality.
  • "Modern" does not require glass, neon, huge cards, tiny text, or springy buttons. Use texture and motion for a stated purpose, not to conceal weak hierarchy.
  • Keep functional text and interaction native. Do not ship a flattened generated screenshot as an interface. Inspect actual alpha, scale, and asset provenance.
  • Focus, hover, selection, equipped, locked, disabled, and busy are not synonyms. Decorative transitions must not silently delay input or commit an action.
  • No blanket ban on manual coordinates, continuous updates, or linear motion. Choose the appropriate coordinate space, update cadence, and semantic behavior.
  • No invented runtime tests, fabricated screenshots, invisible tool capabilities, or claims of accessibility certification. Label unavailable checks explicitly.

Persist decisions and evidence in the project rather than repeating long reference text in chat. For delegation, send the relevant module names, screen contract, asset paths, and acceptance checks—not the whole skill. For installation, tools, validation scripts, and evaluation cases, see README.

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.