agentleFS
BeckhamLabsLLCBeckhamLabsLLC/gemini-visual-design/skills/visual-design-system/SKILL.md

visual-design-system

Skill1 starsChanged 58 days agoMITView on GitHub

This skill provides visual design guidance using Google Gemini's AI capabilities. Use this skill when the user asks to "generate an image", "create a mockup", "design a UI", "make a logo", "create game assets", "generate a video", "review a design", "create icons", "design a landing page", or any visual design/generation task. Also use when writing UI code that could benefit from visual asset generation.

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. KinwiseNZ-DevOps/ui-ux-designer★ 0same 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 repositoryOne other file it gives its agents.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 4 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when working on visual design tasks — image generation, UI mockups, game assets, video creation, or design analysis.

    line 8
  2. Distinctive:

    Never jump straight to expensive generation: Follow this workflow:

    line 12
  3. Distinctive:

    Use negative prompts ("no X", "don't include Y")

    line 37
  4. Very specific:

    Use the pro tier for drafts or iterative exploration (it costs 2x fast)

    line 40
  5. Very specific:

    Every image tier accepts a reference_image, so asking for style consistency never forces a tier change.

    line 55
  6. Very specific:

    Use get_prompt_templates to browse these:

    line 60
  7. Very specific:

    Use analyze_design with the appropriate focus and project_type

    line 70
  8. Distinctive:

    All generation tools automatically apply this profile to prompts.

    line 84

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Visual Design System with Gemini

Use this skill when working on visual design tasks — image generation, UI mockups, game assets, video creation, or design analysis.

Core Workflow: Draft → Edit → Finalize

Never jump straight to expensive generation. Follow this workflow:

  1. Set up style profile first — Use init_style_profile to create a .gemini-design-profile.json if one doesn't exist. This ensures visual consistency across all generations.
  2. Start with a template — Use get_prompt_templates to find a relevant template. Templates provide proven prompt structures that produce consistent results.
  3. Generate a draft — Use generate_image with model="gemini" (or "auto"). This is fast and cheap.
  4. Review and iterate — Use edit_image to refine the draft. Always prefer editing over regenerating from scratch.
  5. Finalize if needed — Only use model="pro" when the user approves a draft and needs production quality.
  6. Save to project — Use save_asset to move from preview to project directory.

Prompt Engineering Rules

DO:

  • Describe scenes narratively — "A warrior standing in a moonlit forest" not "warrior, moon, forest"
  • Use positive framing — "An empty street at dawn" not "A street with no cars"
  • Include composition — Mention shot type, layout, visual hierarchy
  • Specify style explicitly — "Pixel art, 16-bit aesthetic" or "Clean vector illustration"
  • Reference the style profile — Let the profile inject colors, typography, framework context

DON'T:

  • Send raw keyword lists to the API
  • Use negative prompts ("no X", "don't include Y")
  • Skip the style profile when one exists
  • Regenerate from scratch when an edit would work
  • Use the pro tier for drafts or iterative exploration (it costs 2x fast)

Model Selection Guide

Taskmodel=CostWhy
Throwaway iterationdraft~$0.034/imageCheapest; fine for composition checks
Draft mockups, editingfast (default)~$0.067/imageGood quality, multi-turn editing
Final production assetspro~$0.134/imageLegible text, highest fidelity
UnsureautovariesPicks pro for finals, else fast
Design analysis(text model)~$0.001Runs on Gemini Flash text, not an image model
Short video clipsveo-3.1-lite~$0.05/secCheapest video
Balanced videoveo-3.1-fast~$0.10/secDefault
High-quality videoveo-3.1~$0.40/secBest quality

Every image tier accepts a reference_image, so asking for style consistency never forces a tier change.

Template Categories

Use get_prompt_templates to browse these:

  • ui-mockups: Dashboard, settings, login, e-commerce, mobile app
  • game-assets: Characters, textures, icons, backgrounds, UI elements
  • landing-pages: Hero, features, pricing, CTA, testimonials
  • web-components: Navigation, cards, modals, tables, forms
  • icons: App icons, feature sets, illustrations, patterns, logos

Design Analysis Workflow

When reviewing designs:

  1. Use analyze_design with the appropriate focus and project_type
  2. The analysis returns scored categories and specific issues
  3. Each issue includes an edit_instruction — feed it directly to edit_image
  4. Iterate: analyze → edit → analyze again

Style Profile

The .gemini-design-profile.json stores:

  • Color palette (primary, secondary, background, surface, text)
  • Typography preferences (font family, style)
  • Framework context (React, Vue, Tailwind, etc.)
  • Design system (Material, custom, etc.)
  • Visual style keywords

All generation tools automatically apply this profile to prompts.

References

See the reference files for detailed guidance:

  • references/prompt-guide.md — Detailed prompt engineering techniques
  • references/project-types.md — Game, landing page, and web app specific guidance
  • references/prompt-templates.md — Complete template library with examples

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.