Portable Brand Kit Applied Across Every Output
Builds one brand kit file with color roles, contrast-checked pairs, fonts per medium and voice rules, wires it in, and audits drift.
Before & after
Same prompt, same agent — the only difference is the skill.
A brand kit is a good idea for consistency. You'd probably want to define your colors, typography, and maybe your voice guidelines. That way different tools can reference the same rules. The specifics would depend on what brand elements matter most to you.
BRAND KIT SETUP CONFIGURATION
PALETTE DEFINITION
Primary: #0D47A1 (navy)
Accent: #FF6F00 (orange)
Neutral light: #F5F5F5
Neutral dark: #212121
Status - success: #2E7D32
Status - alert: #C62828
TYPOGRAPHY STACK
Headings: Inter SemiBold, 32/28/24/20px
Body: Inter Regular, 16/14px
Mono: Space Mono, 14px (code blocks)
Line height: 1.5 (body), 1.2 (headings)
VOICE RULES
→ Write active voice ("We shipped" not "It was shipped")
→ No Oxford commas in lists (except technical)
→ Use contractions (it's, don't, we've)
→ Avoid exclamation marks; use periods for seriousness
→ When unsure, sound like a colleague explaining something
LOGO ASSETS
• Full logo: 200px min width
• Icon: 48px, on navy or white only
• Lockup (logo + type): 300px min width
OUTPUT INHERITANCE
When installed: All downstream skills (dashboards, decks, landing pages) automatically apply palette, type stack, voice rules. Designers cannot override.
WHITELABEL MODE
For per-client kits: Create brand-kit-clients skill folder with one file per client ID, loaded on demand. Allows different palette/voice per client without re-installing.
ENFORCEMENT LEVEL: Strict (block generation if brand kit conflicts detected)About this skill
Your deck uses one blue, your landing page uses another, and the chart Claude made yesterday uses a third. You wrote "friendly but professional" somewhere, and nothing follows it. This skill turns your brand into one kit file that sits in context Claude always loads, so every output draws from the same values.
What it does
- Gives every color a role (action, text, surface, highlight) so outputs stop choosing brand colors at random
- Checks the contrast ratio of each text-and-background pair and creates a darker text-safe version when a brand color fails
- Defines font stacks for web, Office docs, and email, so a missing font falls back to something you approved instead of Times New Roman
- Turns voice adjectives into rules you can check: sentence length, punctuation, banned words with replacements, and before/after rewrites
- Sets up one kit per client and shows how to switch between them without one client's brand leaking into another's work
- Audits pasted outputs for off-brand hex codes, near-miss colors, wrong fonts, and voice violations, with a fix list
What you get
A versioned BRAND.md with color tokens, approved contrast pairs, chart colors, type rules, voice rules, logo rules, and CSS/Tailwind/Python exports. You also get the snippet that loads it into your project, and a list of brand problems to fix.
Best for
Agencies running several client brands through Claude, and in-house marketers whose decks, pages, docs, and emails don't look like the same company.
SKILL.md preview
---
name: portable-brand-kit-across-outputs
… (sign up to view the full skill)Featured in
More marketing & growth skills
View all Marketing & Growth skills →Paid Ads Campaign Planning and ROAS/CPA Diagnosis
Plans paid campaigns on Google, Meta, LinkedIn and X from break-even math, writes ads, and finds the metric behind a ROAS or CPA drop.
Scheduled Marketing Loop Spec for AI Agents
Designs a recurring marketing loop an AI agent runs on schedule: fresh inputs, ledger, approval gates, metrics, kill rules, run prompt.
App Store Screenshot Headlines That Convert
Screenshot captions for App Store and Google Play, mined from your app's code: ordered headlines, alternates, and A/B test ideas.
Analytics Audit: Reconcile Numbers, Fix Tracking
Finds which number is true when your dashboards disagree, sizes every gap, and builds tracking around the decision you have to make.