agentleFS
vidual-labsvidual-labs/digital-marketing-skills/skills/marketing/branding/SKILL.md

branding

Skill3 starsChanged 45 days agoMITView on GitHub

Use when defining, refreshing or documenting a brand identity system — visual identity (logo, colour, typography, imagery, layout), brand voice and tone matrix, vocabulary, copy standards by channel, naming conventions, touchpoint consistency and brand guidelines for humans and AI content tools. Also use when someone asks for "brand guidelines", "tone of voice", "a style guide for our content", "our brand feels inconsistent" or needs a brand brief for designers or AI writing. Don't use for one-off ad copy (meta-ads-creative, tiktok-ads), competitive positioning (market-positioning) or competitor analysis (competitor-research).

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 2

    Distinctive

    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. Very specific:

    Don't use for: single ad or post copy (use meta-ads-creative, tiktok-ads), competitive positioning and messaging architecture (use …

    line 30
  2. Distinctive:

    Ask for (or extract from the conversation):

    line 34
  3. Distinctive:

    Test: shuffle brand images with competitors' — if yours are not identifiable at a glance, the style is not distinctive enough.

    line 120

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “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

Branding

Overview

Build a cohesive brand identity system — visual, verbal, strategic and experiential — so every output is recognisable before the logo is seen. Covers visual identity, voice and tone, vocabulary, copy standards per channel, naming, touchpoint consistency and the practical guidelines that keep humans, agencies and AI tools producing the same brand. A brand is the pattern of associations an audience builds across touchpoints; this skill makes that pattern deliberate.

When to Use

  • Creating a brand identity from scratch (startup, product, market entry)
  • Refreshing a brand that feels stale, inconsistent or dated
  • Writing brand guidelines for a team, agency, freelancers or AI content workflows
  • Defining voice and tone before a content or campaign programme
  • Naming a product, feature or sub-brand
  • Auditing touchpoints for consistency

Don't use for: single ad or post copy (use meta-ads-creative, tiktok-ads), competitive positioning and messaging architecture (use market-positioning), or competitor analysis (use competitor-research).

Inputs

Ask for (or extract from the conversation):

  • Brand or working name, what it sells, to whom, and the price tier
  • Positioning statement or point of difference (run market-positioning first if none exists; otherwise capture a one-line hypothesis)
  • Existing assets: logo files, colours, fonts, templates, past guidelines
  • Audience and context: where the brand is encountered (web, app, social, retail, print, packaging)
  • Competitor look and feel to differentiate from
  • Constraints: accessibility requirements, legal (trademarks, regulated claims), languages and scripts, budget for fonts and photography
  • Personality cues: 3–5 adjectives the team uses, and 3 they reject

Data rules: do not invent hex codes for an existing brand — ask for them or mark values as proposals. Check trademarks and domains only if you can; otherwise flag them as to-do items. Where this skill cites standards (WCAG levels, platform safe zones) the versions are current as of metadata.updated; confirm against the platform if it matters legally.

Brand System Architecture

PillarComponentsDeliverable
StrategicPositioning, promise, values, personalityBrand foundation (one page)
VerbalVoice attributes, tone matrix, vocabulary, naming, taglineVoice & tone guide
VisualLogo, colour, typography, imagery, iconography, layout, motionVisual identity guide
ExperientialTouchpoint standards, templates, content and AI usage rulesTouchpoint playbook

Build in that order. Visual decisions made before the strategic and verbal ones get redone.

1. Strategic Foundation

  • Promise: one sentence of what the audience can always expect
  • Personality: 3–4 adjectives with "is / is not" boundaries (used again in voice)
  • Values: 3–5, each with a behaviour that proves it
  • Positioning: from market-positioning — the frame of reference and point of difference the identity must express

2. Visual Identity

2.1 Logo

DecisionGuidance
TypeWordmark, lettermark, pictorial, abstract or combination; choose for memorability and scalability
VariantsPrimary, horizontal, stacked, icon/lettermark, monochrome, reversed
SizesLegible at 16 px (favicon), 48–64 px (avatars), 500 px (web), large format
Clear spaceAt least the height of the main letterform around the logo
Don'tsNo stretching, recolouring outside the palette, effects, or placing on busy imagery
FilesSVG master, PNG with transparency, dark and light versions

2.2 Colour

RoleCountPurpose
Primary1The colour the brand owns: logo, primary actions
Secondary2–3Hierarchy, sections, illustration
Accent1–2Highlights and interactive states
Neutrals4–6Text, backgrounds, borders, surfaces (define light and dark mode sets)
Semantic4Success, warning, error, info — must not clash with brand colours

Rules: exact hex (and optionally OKLCH/HSL for design systems); every text/background pairing meets WCAG 2.2 AA (4.5:1 body, 3:1 large text and UI components); check the palette in greyscale for hierarchy; define which pairings are allowed; never rely on colour alone to convey meaning.

2.3 Typography

ElementSpecification
Primary typefaceOne family for headings and body; variable fonts simplify weights
SecondaryOptional (display or monospace) only with a defined job
ScaleNamed steps (display, h1–h4, body, small, caption) on a ratio (1.2–1.333) with rem values
Weights2–4 maximum
Line heightBody 1.5–1.7; headings 1.1–1.25
FallbacksSystem stack per platform; test with fallbacks active
LicensingConfirm web, app and desktop licences; open-source fonts avoid the problem
PersonalityTypeface directionOpen-source examples
Modern, cleanGeometric sansInter, Manrope, Outfit
Trustworthy, humanHumanist sansSource Sans 3, Nunito Sans, Figtree
Playful, creativeRounded or displayFredoka, Quicksand, Bricolage Grotesque
Premium, editorialSerifPlayfair Display, Lora, Fraunces
TechnicalMonospace / groteskJetBrains Mono, IBM Plex Mono, Space Grotesk

2.4 Imagery and illustration

ElementDecide
PhotographyLifestyle, product, documentary, abstract; lighting, colour grading, composition rules
IllustrationFlat, line, 3D, hand-drawn, or none; stroke, palette, perspective
PeopleRepresentation guidelines, candid vs posed, real customers vs stock
Screenshots / UIDevice frames, shadows, background treatment
AI-generated imageryAllowed or not; required disclosure; style prompts and a do-not list; licence and model policy
Alt textPart of the voice; descriptive, no "image of"

Test: shuffle brand images with competitors' — if yours are not identifiable at a glance, the style is not distinctive enough.

2.5 Layout, spacing, motion

ElementGuideline
Grid12-column web, 4-column mobile, consistent gutters
Spacing scale4/8 px base (4, 8, 12, 16, 24, 32, 48, 64)
RadiusOne system (0 / 4 / 8 / 16 px / pill)
Elevation2–3 shadow levels or none
Max width~65–75 characters per line for reading; 1200–1440 px layouts
MotionDurations (150–300 ms UI), easing, what animates and what never does; respect reduced-motion settings
Dark modePalette mapping defined, not inverted automatically

3. Brand Voice

3.1 Voice attributes (constant)

Define 3–4 attributes, each with what it means and what it is not:

Attribute (example)MeansIs not
Plain-spokenShort words, concrete nouns, no jargonDumbed down, slangy
ExpertLeads with evidence and specificsAcademic, condescending
WarmAddresses the reader, acknowledges effortGushing, emoji-heavy
DirectSays the point firstBlunt, dismissive

"Friendly" and "professional" alone are table stakes, not attributes; the attribute is how the brand is friendly differently.

3.2 Tone matrix (adapts by context)

ContextToneExample shift
Launch / announcementEnergetic, confidentBenefit-led headline, active verbs
Error / outageCalm, specific, accountableWhat happened, what to do, when it is fixed
Education / helpPatient, step-by-stepNumbered steps, no assumed knowledge
Sales / conversionClear, evidence-ledOutcome, proof, one next step
Support replyEmpathetic, solution-firstAcknowledge, resolve, confirm
Social replyConversational, briefHuman, in-joke aware, never corporate
Legal / policyPlain, preciseShort sentences, defined terms
CrisisTransparent, measuredFacts, responsibility, next update time

3.3 Vocabulary

A use / avoid list makes writing recognisable and keeps product terms consistent.

UseAvoidWhy
"Free trial""Complimentary period"One name for one offer
"Get started""Begin your journey"Cliché
"You / your""Users / customers" (in copy)Direct address
Product term listSynonyms for the same featureSearchability and clarity

Minimum 10 pairs; 25+ for teams. Add spelling conventions (sentence case, serial comma, numerals, date and currency formats, localisation notes).

3.4 Sentence style

Average 12–20 words; vary rhythm; active voice; one idea per paragraph; headlines lead with the benefit or action; ban list of filler ("leverage", "seamless", "cutting-edge", "we're excited to announce"); prefer specific numbers and named examples; inclusive language by default.

4. Copy Standards by Channel

ChannelLengthPatternGoal
Homepage headline5–10 wordsOutcome for whomCommunicate position in 3 seconds
Ad headline≤ 27–40 characters (Meta) / 30 (Google RSA)Offer or outcomeStop the scroll
Ad primary text≤ 125 characters before the foldHook → proof → CTAClick
Landing page body200–600 wordsProblem → proof → solution → steps → proof → objections → CTAConvert
Social post1–3 short paragraphs; first line carries the ideaNative to each platformEngagement, saves, sends
Email subject3–7 wordsSpecific benefit or curiosity, no baitOpen
Email body80–250 wordsOne idea, one CTAClick
Product description50–150 wordsFeature → benefit → specDecide
Help articleAs long as needed, scannableAnswer first, steps, screenshotsResolve
Alt text≤ 125 charactersDescribe content and functionAccessibility

CTA rules

One primary CTA per screen; verb first ("Get", "Start", "Book"); specific ("Get the pricing sheet" beats "Download"); primary CTA uses the accent/action colour, secondary is outline or text; minimum 44×44 px tap target.

5. Naming Conventions

AssetConvention
Products / sub-brandsPronounceable, memorable, available (domain, handles, trademark search); max 3 syllables preferred; architecture decided (branded house vs house of brands)
FeaturesFunctional and benefit-oriented ("Smart Search"); avoid invented words users cannot search
CampaignsInternal YYYY-channel-campaign; external emotional or benefit-led
Filesbrand_channel_asset_version_YYYY-MM-DD.ext
Design tokenscolor.brand.primary, space.4, font.heading

6. Touchpoint Consistency and AI Usage

TouchpointCheckStandard
Website / appColours, type, imagery, voice, CTADesign tokens in code
Social profilesAvatar, banner, bio voice, pinned contentTemplate per platform with safe zones
AdsPalette, type, copy voice, imagery styleAd template library
EmailHeader, signature, voice, button styleMaster template
Decks and documentsTemplate, charts, typeMaster templates
Help and docsVoice, formatting, screenshotsWriting guide for contributors
Packaging / physicalLogo placement, colour matching (Pantone/CMYK), typePrint guide
AI content toolsSystem prompt with voice, vocabulary, banned words, examples; review step before publishing; disclosure policyBrand prompt pack

The brand prompt pack (voice attributes, tone matrix, vocabulary, three good and three bad examples, visual style prompts) is now a standard deliverable: it is how the brand survives AI-assisted production.

Output Format

BRAND IDENTITY GUIDE: [Brand]
Date: [Date] | Status: [proposal / approved] | Positioning: [one line]

--- STRATEGIC FOUNDATION ---
Promise: [..] | Personality: [adj (is / is not) ×3–4] | Values: [value → behaviour]

--- VISUAL IDENTITY ---
Logo: [type, variants, clear space, files]
Colour: Primary [hex] | Secondary [hex…] | Accent [hex] | Neutrals [hex…] | Semantic [hex…]
  Accessibility: [pairings checked, WCAG 2.2 AA pass/fail]
Typography: [family, weights, scale with ratio, fallbacks, licence]
Imagery: [photo, illustration, people, screenshots, AI policy, alt text]
Layout & motion: [grid, spacing, radius, elevation, max width, motion, dark mode]

--- BRAND VOICE ---
Attributes: [attribute — means — is not] ×3–4
Tone matrix: [context → tone → example] ×6+
Vocabulary: Use [..] / Avoid [..] (10+ pairs) | Conventions: [case, numerals, dates, currency]
Sentence style: [rules]

--- COPY STANDARDS ---
[Channel → length → pattern → goal] for the channels the brand uses
CTA rules: [..]

--- NAMING ---
[Products, features, campaigns, files, tokens]

--- TOUCHPOINT AUDIT ---
[Touchpoint → pass/fail → fix] ×5+

--- AI USAGE ---
Brand prompt pack: [voice, vocabulary, examples, visual prompts, review and disclosure rules]

--- ASSET CHECKLIST ---
[ ] Logo variants (SVG/PNG, dark/light)  [ ] Colour tokens  [ ] Font files/licences
[ ] Image references (3–5)  [ ] Social templates with safe zones  [ ] Email template
[ ] Deck template  [ ] Writing guide  [ ] Brand prompt pack  [ ] Guide distributed

--- OPEN ITEMS ---
[Trademark/domain checks, approvals, missing assets]

Common Pitfalls

  1. Starting with the logo. Strategy and voice first; the logo expresses them.
  2. Too many colours. One hero colour; everything else supports it.
  3. Vague voice words. "Friendly, professional" describe every brand. Define the how and the is not.
  4. Tone drift between channels. Tone adapts, voice does not; a brand that is playful on Instagram and stiff on LinkedIn reads as two brands.
  5. Unchecked accessibility. Pretty pairings that fail 4.5:1 exclude readers and fail audits. Check every pairing.
  6. Fonts nobody can use. Confirm licences for web, app and desktop before specifying; open-source fonts avoid surprises.
  7. A 60-page guide nobody opens. Lead with examples; keep the rules scannable; publish tokens and templates, not just PDFs.
  8. No rules for AI tools. Without a prompt pack and a review step, AI output drifts off-brand fastest of all.

Verification Checklist

  • Strategic foundation stated (promise, personality with is/is-not, values, positioning line)
  • Colour palette with exact hex for primary, secondary, accent, neutrals and semantic colours
  • Text/background pairings checked against WCAG 2.2 AA with pass/fail noted
  • Typography specified with weights, scale, fallbacks and licence status
  • Imagery rules cover photography, illustration, people, screenshots, AI-generated media and alt text
  • Layout, spacing, radius, motion and dark-mode rules defined
  • 3–4 voice attributes each with "means" and "is not"
  • Tone matrix covers at least 6 contexts with examples
  • Vocabulary list has 10+ use/avoid pairs plus writing conventions
  • Copy standards and CTA rules given for the channels the brand uses
  • Naming conventions for products, features, campaigns, files and tokens
  • Touchpoint audit of 5+ touchpoints with pass/fail and fixes
  • Brand prompt pack for AI tools included
  • Existing brand values not invented; proposals labelled as proposals; open items listed

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.