How do I install the Branding skill?

Run npx skillmds@latest add steveclarke/branding in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Branding skill do?

Create a complete brand identity — logo authored directly as SVG, colors, fonts, SVG refinement in Inkscape, and a full asset pack (PNGs, favicons, OG images), delivered as a brand guide HTML page. Triggers on branding, logo, brand pack, brand color, project identity, social preview, visual identity. It is listed under Design & Media on SkillMD.

Is Branding safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Branding?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Branding free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Branding?

steveclarke (@steveclarke) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Branding by steveclarke — Agent Skill

Name: Branding
Author: steveclarke

Branding

Create a complete brand identity for a project: logo mark, wordmark lockups, color palette, font selection, a production-ready asset pack, and a brand guide.

Prerequisites

Verify these tools are installed before starting:

which inkscape && which magick

Inkscape — SVG manipulation, text-to-path conversion, PNG export
- macOS: brew install --cask inkscape
ImageMagick v7+ — image compositing, resizing, format conversion
- macOS: brew install imagemagick

Overview

The workflow has 7 phases. The first 3 are collaborative with the user; the last 4 are mostly automated production work.

Discovery — understand the project and what the logo should convey
Logo generation — author SVG concepts directly and review them with the user
Font and color selection — pick a wordmark font and brand color
SVG refinement — Inkscape: text to paths, viewBox cleanup, lockup creation
Asset pack production — export all variants, sizes, and formats
Brand guide — generate a comprehensive HTML brand reference page
Project wiring — place assets where they're needed (optional)

Browser-Based Previews

Throughout the collaborative phases, use temporary HTML pages to show the user their options. This is a core technique in this workflow — never ask the user to open SVG files from folders or guess what things look like.

The pattern:

Generate a self-contained HTML file in the project's tmp/ directory (e.g., tmp/brand-fonts.html)
Open it in Chrome: open -a "Google Chrome" /path/to/file.html
The user reviews and gives feedback
Clean up tmp/ files before committing

Key rules for preview pages:

Use plain HTML and CSS only — avoid JavaScript where possible. If JS is needed, use safe DOM methods (textContent, createElement) rather than innerHTML, which security hooks may block.
Load Google Fonts via <link> tags in the <head> for font previews — they render instantly in the browser.
Inline SVG marks directly in the HTML for lockup previews (no external file references).
Keep pages self-contained — no external dependencies beyond Google Fonts CDN.

Phase 1: Discovery

Ask the user:

Project name — exact casing and spelling for the wordmark
What it does — one sentence, so you understand the domain
Logo vibe — abstract geometric? lettermark? mascot? isometric? minimal?
Any symbols or metaphors — things the name evokes (e.g., "Kiso" means "foundation" in Japanese, which led to stacked building blocks)
Color preferences — warm/cool, any existing colors, or "surprise me"

Use answers already available in the conversation. Keep discovery brief, then create concepts.

Phase 2: Logo Generation

Author the logo directly as editable SVG using the latest available flagship model from OpenAI or Anthropic, with the highest supported reasoning or effort setting. Prefer the strongest available model for visual design and code generation; do not hard-code model names or versions in this skill. Respect the user's explicit model choice. If the environment cannot change models or effort, state that briefly and work with the available model without blocking progress or claiming a switch occurred.

Use the discovery brief to create a few distinct concepts when the user wants options, or refine the chosen direction directly. Write real vector geometry (paths, shapes, fills, and strokes), with a clean viewBox and no embedded raster image standing in for the logo. Keep the mark separate from the wordmark, which gets a proper font in Phase 3.

Render the SVGs and inspect them visually before presenting them. Show concepts together on an HTML preview page, including small icon sizes and light/dark backgrounds. Check the silhouette, spacing, stroke consistency, and legibility. Iterate from the user's feedback; keep the approved version intact while exploring alternatives and replace it only when the user chooses a successor.

Third-party SVG generators such as quiver.ai or recraft.ai are optional fallbacks when the user requests one or direct SVG iteration cannot achieve the desired result. Do not require accounts, external generation, or a copy-and-paste prompt handoff as the normal workflow.

Once a mark is chosen, inspect its SVG structure (viewBox, dimensions, colors, and any embedded text or fonts) before producing the asset pack. Convert any remaining text to paths in Phase 4.

Phase 3: Font and Color Selection

Font

Suggest 4-6 fonts that pair well with the mark's style. Build a single HTML preview page that shows all candidates side by side:

Each font rendered as the project name in bold (700/800) at large size using the proposed brand color
A second line in regular weight (400) for comparison
A body text sample ("The quick brown fox...")
Weight demos (400, 600, 700, 800) if the font supports them
Brief notes on personality and fit (e.g., "Geometric, modern — closest to current wordmark")

Load fonts via Google Fonts <link> tags. Open the page in Chrome for the user to compare.

After the user picks a font:

Download from the Google Fonts GitHub repo (most reliable source):
```
curl -L -o /tmp/FontName.ttf "https://raw.githubusercontent.com/google/fonts/main/ofl/fontname/FontName%5Bwght%5D.ttf"
```
Note: use the ofl/ path for most Google Fonts. Verify the download with file /tmp/FontName.ttf — it should say "TrueType Font data", not "HTML" or "ASCII text".
Install locally: cp /tmp/FontName.ttf ~/Library/Fonts/ (macOS)
Save to brand/fonts/ for the asset pack

Color

If the user already has a logo with colors, extract them first. Show all colors found in the SVG as chips in a preview HTML page — name each one by its role (e.g., "Hull Navy", "Sail Red").

Then present brand color candidates as large clickable-looking cards with:

A big color swatch showing the hex
A name and description of its personality
A recommendation note on the best one

Include a proposed color schema section showing how the colors would be used:

Primary — buttons, links, active states
Dark/Navy — sidebar, headings, dark surfaces
Accent — alerts, destructive actions, visual interest
Highlight — info banners, selected states

Color variant exploration: If the user wants to try different shades or swap a color (e.g., "that red is too dark"), generate a comparison page with 4-6 variants showing the full mark with each color swapped in. Use sed to replace hex values in the SVG for quick iteration — inline each variant SVG directly in the HTML.

Tailwind color scale: When the user settles on a brand color, they'll need a full shade scale for Tailwind. Direct them to generate one at:

uicolors.app — excellent OKLCH-based scales
kigen.design/color — alternative generator

The user pastes their hex, gets back a 50-950 scale in OKLCH, and shares it with you to wire into the project.

Recording color decisions

Record the final colors for the asset pack:

Primary: the main brand color with its Tailwind scale name
All logo colors: hex values and roles (hull, sails, water, etc.)
Tailwind scale: the full 50-950 OKLCH values

Phase 4: SVG Refinement

This phase produces the master SVGs that everything else is derived from. You need 9 SVG files:

Variant	Mark only	Horizontal lockup	Stacked lockup
Color	mark-color.svg	logo-horizontal-color.svg	logo-stacked-color.svg
Black	mark-black.svg	logo-horizontal-black.svg	logo-stacked-black.svg
White	mark-white.svg	logo-horizontal-white.svg	logo-stacked-white.svg

Creating lockups

A lockup is the mark + wordmark text composed together. Two layouts:

Horizontal — mark on the left, text to the right, vertically centered
Stacked — mark on top, text centered below

Start with the color horizontal lockup:

Open the mark SVG, note its dimensions
Create a new SVG combining the mark with a <text> element using the chosen font
Size the text proportionally — the text cap-height should be roughly 40-60% of the mark height for horizontal, and the text width should roughly match the mark width for stacked
Set the viewBox to tightly fit the content with minimal padding

Preview each lockup in Chrome before converting text to paths. Open the SVG directly — open -a "Google Chrome" file.svg. The user should confirm the proportions, spacing, and overall feel before proceeding. Adjust text x position, font-size, and viewBox width based on feedback.

Text to paths

Google Fonts @import doesn't work in Inkscape. The font must be installed locally first (done in Phase 3). Then convert all text to paths so the SVG is self-contained:

inkscape input.svg --actions="select-all;object-to-path;export-filename:output.svg;export-do"

After conversion, verify with grep -c '<text' output.svg — should return 0.

Color variants

Black: replace all fill colors with #18181b (zinc-950, not pure black — standard design practice for better readability)
White: replace all fill colors with #ffffff
For lockups, the mark keeps its colors in the "color" variant; text is #18181b. In black/white variants, everything is monochrome.

Use sed for color replacement — handle both fill="#hex" attributes and fill:#hex in style attributes:

sed 's/fill="#[0-9A-Fa-f]\{6\}"/fill="#18181b"/g; s/fill:#[0-9a-fA-F]\{6\}/fill:#18181b/g' input.svg > output.svg

ViewBox cleanup

After compositing, trim the viewBox to fit content tightly:

inkscape file.svg --actions="select-all;fit-canvas-to-selection;export-filename:file.svg;export-do"

Or manually adjust the viewBox after checking element bounding boxes:

inkscape file.svg --query-all

Phase 5: Asset Pack Production

Directory structure

brand/
├── svg/          9 master SVGs (from Phase 4)
├── png/          rasterized variants at standard sizes
├── favicon/      ico + png sizes + webmanifest
├── social/       platform-specific images
└── fonts/        font files + README

PNG exports

Export from the master SVGs using Inkscape (better SVG rendering than ImageMagick):

Marks — export at: 64, 128, 256, 512, 1024px (square, width = height)

inkscape mark-color.svg --export-type=png --export-filename=../png/mark-color-256.png --export-width=256

Lockups — export at widths appropriate for the aspect ratio:

Horizontal: 400, 600, 800px wide
Stacked: 300, 500px wide

Favicons

Use the /favicon skill if available. Otherwise, generate manually from the mark:

# PNG sizes for favicons
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/favicon/favicon-16x16.png --export-width=16
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/favicon/favicon-32x32.png --export-width=32
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/favicon/apple-touch-icon.png --export-width=180
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/favicon/android-chrome-192x192.png --export-width=192
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/favicon/android-chrome-512x512.png --export-width=512

# ICO (multi-resolution)
magick brand/favicon/favicon-16x16.png brand/favicon/favicon-32x32.png brand/favicon/favicon.ico

Create brand/favicon/site.webmanifest:

{
  "name": "PROJECT_NAME",
  "short_name": "PROJECT_NAME",
  "icons": [
    { "src": "/android-chrome-192x192.png", "sizes": "192x192", "type": "image/png" },
    { "src": "/android-chrome-512x512.png", "sizes": "512x512", "type": "image/png" }
  ],
  "theme_color": "#HEX_OF_BRAND_COLOR",
  "background_color": "#ffffff",
  "display": "standalone"
}

Social images

OG image (1280x640) — for GitHub social preview, Open Graph tags:

magick -size 1280x640 xc:white \
  \( brand/svg/logo-horizontal-color.svg -resize 1100x -trim +repage \) \
  -gravity center -composite \
  brand/social/og-image-1280x640.png

The -trim +repage before -gravity center is critical — it strips the SVG's internal whitespace so ImageMagick centers the actual visual content, not the viewBox bounding box. Without this, the logo will appear off-center.

Platform icons:

Discord server icon: 512x512 from the mark
GitHub avatar: 500x500 from the mark
npm icon: 256x256 from the mark

inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/social/discord-icon-512.png --export-width=512
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/social/github-avatar-500.png --export-width=500
inkscape brand/svg/mark-color.svg --export-type=png --export-filename=brand/social/npm-icon-256.png --export-width=256

Fonts README

Create brand/fonts/README.md documenting the font choice, license, and source:

# Brand Font

**[Font Name]** by [Foundry]

- Weight: 700 (Bold) for wordmark
- License: SIL Open Font License
- Source: [GitHub URL]

The full font family archive is included for reference.

Phase 6: Brand Guide

Generate a comprehensive HTML brand guide page at brand/brand-guide.html. This is what a designer would deliver as a "brand style guide" or "brand board" — a single-page reference that anyone on the team can open to understand how to use the brand.

The brand guide should include these sections:

1. Header

Project name and tagline/description at the top.

2. Logo Versions

Show all 9 SVG variants organized in a grid:

Color row: mark, horizontal lockup, stacked lockup (on white background)
Black row: same three (on white background)
White row: same three (on dark background, e.g., #0f172a)

Label each with its filename and recommended usage:

Mark — favicons, app icons, social avatars, small spaces
Horizontal — headers, README badges, email signatures
Stacked — splash screens, print, when vertical space allows

Include minimum size guidance: mark should not be displayed smaller than 32px, lockups not smaller than 120px wide.

3. Color Palette

Show each brand color as a large swatch with:

Color name and role (e.g., "Primary — buttons, links, focus rings")
Hex value
OKLCH value (if available)
Tailwind class (e.g., bg-primary-500)

If a full Tailwind scale was generated, show the complete 50-950 ramp as a horizontal strip of swatches with labels.

Show all logo-specific colors too (hull, sails, water, etc.) with their hex values.

4. Typography

Font name, foundry, license
Wordmark weight and letter-spacing
Preview of the font at the wordmark weight, plus regular weight
Note the font source (Google Fonts, GitHub repo) and where files are stored in brand/fonts/

5. Asset Inventory

A table listing every file in the brand/ directory with:

Filename
Dimensions or purpose
Format

Organized by subdirectory (svg/, png/, favicon/, social/, fonts/).

Construction notes

Self-contained HTML — inline all SVGs directly, load the brand font via Google Fonts <link>. No external file references (the guide should work when opened from any location).
No JavaScript — pure HTML and CSS. This ensures it works everywhere and avoids security hook issues.
Clean, professional design — use the brand colors themselves to style the page. White background, brand-colored headings and accents.
Print-friendly — reasonable layout at standard page widths.

After generating, open in Chrome for the user to review: open -a "Google Chrome" brand/brand-guide.html

Phase 7: Project Wiring (Optional)

This phase is project-specific. Ask the user which of these apply:

README

<p align="center">
  <a href="PROJECT_URL">
    <img src="brand/svg/logo-horizontal-color.svg" width="280" alt="PROJECT_NAME">
  </a>
</p>

GitHub renders SVGs from the repo inline. The width attribute prevents it from blowing up to full width. Wrap in <a> to link to the project's website.

Docs site / web app

Copy favicons to the static assets directory
Add favicon <link> tags to <head>
Add <meta name="theme-color" content="#HEX">
Add mark to the navbar/header

Tailwind theme

If the user has a Tailwind scale, wire it into the CSS:

@theme {
  --color-primary-50: oklch(...);
  --color-primary-100: oklch(...);
  /* ... */
  --color-primary-950: oklch(...);
}

GitHub

Social preview: repo Settings → Social preview → upload OG image (manual step)
Avatar: only applies if using a GitHub Organization (manual step)

npm

Add "homepage" to package.json pointing to the docs/website

Other platforms

Discord server icon (manual upload)
1Password: note which account owns the Discord server

Present manual steps clearly at the end — don't try to automate things that require browser UI interaction.

Checklist

Before declaring done, verify:

9 master SVGs with text converted to paths (no font dependencies)
PNG exports at all standard sizes
Favicon set (ico, 16, 32, 180, 192, 512, webmanifest)
OG image at 1280x640, visually centered
Platform icons (Discord 512, GitHub 500, npm 256)
Font files saved with README
Brand guide HTML generated and reviewed
All assets committed to brand/ directory
Temporary preview files in tmp/ cleaned up
Project wiring completed (if applicable)
Manual steps communicated to user

Branding

File contents

Branding

Prerequisites

Overview

Browser-Based Previews

Phase 1: Discovery

Phase 2: Logo Generation

Phase 3: Font and Color Selection

Font

Color

Recording color decisions

Phase 4: SVG Refinement

Creating lockups

Text to paths

Color variants

ViewBox cleanup

Phase 5: Asset Pack Production

Directory structure

PNG exports

Favicons

Social images

Fonts README

Phase 6: Brand Guide

1. Header

2. Logo Versions

3. Color Palette

4. Typography

5. Asset Inventory

Construction notes

Phase 7: Project Wiring (Optional)

README

Docs site / web app

Tailwind theme

GitHub

npm

Other platforms

Checklist

Frequently asked questions

Branding

File contents

Branding

Prerequisites

Overview

Browser-Based Previews

Phase 1: Discovery

Phase 2: Logo Generation

Phase 3: Font and Color Selection

Font

Color

Recording color decisions

Phase 4: SVG Refinement

Creating lockups

Text to paths

Color variants

ViewBox cleanup

Phase 5: Asset Pack Production

Directory structure

PNG exports

Favicons

Social images

Fonts README

Phase 6: Brand Guide

1. Header

2. Logo Versions

3. Color Palette

4. Typography

5. Asset Inventory

Construction notes

Phase 7: Project Wiring (Optional)

README

Docs site / web app

Tailwind theme

GitHub

npm

Other platforms

Checklist

Related skills

Frequently asked questions

How do I install the Branding skill?

What does the Branding skill do?

Is Branding safe to use?

Which AI agents work with Branding?

Is Branding free to use?

Who published Branding?