Skip to content
cccskills
FreePublished on GitHub

brand-voice-learner

Analyze existing brand content to extract voice patterns, create voice guidelines, and ensure consistent brand expression

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150 stars(GitHub 150 + cccskills 0)

Run score Not tested

See how to install

Files included (1)

  • SKILL.md9.5 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Brand Voice Learner

Analyze existing brand content to extract voice characteristics, create actionable voice guidelines, and ensure consistent brand expression across all communications.

When to Use This Skill

  • Creating brand voice guidelines
  • Onboarding new writers
  • Auditing voice consistency
  • Adapting voice for channels
  • Training AI writing tools

Methodology Foundation

Based on NN/g voice research and content strategy best practices, combining:

  • Linguistic pattern analysis
  • Voice dimension mapping
  • Guidelines development
  • Consistency measurement

What Claude Does vs What You Decide

Claude DoesYou Decide
Analyzes existing contentContent sources to analyze
Extracts voice patternsVoice evolution direction
Creates voice guidelinesApproval of guidelines
Identifies inconsistenciesException handling
Suggests voice examplesFinal voice choices

Instructions

Step 1: Gather Voice Samples

Content to Analyze:

SourcePurpose
Website copyCore brand voice
Blog postsExtended voice
Social mediaCasual variation
Email campaignsDirect communication
Product UIFunctional voice
Customer supportEmpathetic voice

Sample Requirements:

  • Minimum 10-15 pieces
  • Variety of contexts
  • Include best examples
  • Cover different tones

Step 2: Analyze Voice Dimensions

Voice Dimension Framework:

DimensionSpectrum
FormalityCasual ←→ Formal
PersonalitySerious ←→ Playful
DirectnessIndirect ←→ Direct
EnthusiasmReserved ←→ Enthusiastic
TechnicalSimple ←→ Technical
WarmthDistant ←→ Warm

Step 3: Extract Patterns

Linguistic Elements:

ElementWhat to Identify
Sentence lengthShort, medium, long
Word choiceCommon, specialized, branded
ContractionsUsed, avoided, mixed
Active/passivePreference
PronounsWe/you/they patterns
PunctuationExclamations, ellipses, etc.

Step 4: Document Guidelines

Guidelines Structure:

  1. Voice principles (3-5)
  2. Dimension scores
  3. Do/Don't examples
  4. Channel variations
  5. Word lists

Step 5: Create Consistency Tools

Tools to Create:

  • Voice checklist
  • Example library
  • Word/phrase dictionary
  • Channel guidelines

Examples

Example 1: Complete Voice Analysis

Input:

Analyze brand voice from these samples:

Website: "We build tools that actually work. No fluff, no complexity—
just software that helps you get stuff done."

Blog: "Here's the thing about productivity apps: most of them make
your life harder, not easier. We think that's backwards."

Email: "Hey there! Your trial is almost over. Before you go, we wanted
to make sure you saw everything [Product] can do."

Social: "Mondays are hard. Your software shouldn't be. ☕"

Output:

## Brand Voice Analysis

### Voice Summary

**Overall Character:** Confident, casual, anti-corporate

Your brand sounds like a smart friend who's direct and relatable—
someone who gets frustrated by the same things your customers do
and isn't afraid to call out industry BS.

---

### Voice Dimensions

Formality [===|======] Casual (3/10) Personality [======|===] Playful (7/10) Directness [========|=] Very Direct (9/10) Enthusiasm [=====|====] Moderate (6/10) Technical [====|=====] Accessible (4/10) Warmth [======|===] Warm (7/10)


---

### Voice Principles

#### 1. Say it Straight
No corporate speak, no jargon inflation. If you can say it
in fewer words, do it.

**Examples:**
- ✅ "We build tools that actually work"
- ❌ "We leverage cutting-edge solutions to optimize..."

#### 2. Take Their Side
Position as ally against frustration. You share their pain
with the status quo.

**Examples:**
- ✅ "Most productivity apps make your life harder"
- ❌ "Our competitors offer suboptimal experiences"

#### 3. Stay Human
Contractions, casual language, occasional humor. Sound like
a person, not a press release.

**Examples:**
- ✅ "Here's the thing about..."
- ❌ "It is important to note that..."

#### 4. Confident, Not Arrogant
State opinions clearly without hedging, but don't put
others down.

**Examples:**
- ✅ "No fluff, no complexity"
- ❌ "Unlike those other terrible apps..."

---

### Linguistic Patterns

| Element | Pattern | Examples |
|---------|---------|----------|
| Sentence length | Short to medium (8-15 words average) | "We think that's backwards." |
| Contractions | Always used | "That's", "shouldn't", "we're" |
| Pronouns | Heavy "we/you" | Creates conversational feel |
| Active voice | Strong preference | "We build" not "Tools are built" |
| Punctuation | Occasional em-dashes, rare exclamations | "No fluff—just software" |
| Emoji | Occasional, relevant | Coffee emoji for Monday post |

---

### Voice Do's and Don'ts

#### DO ✅

| Pattern | Example |
|---------|---------|
| Use contractions | "We're building" not "We are building" |
| Start with "Here's the thing" | Conversational opener |
| Use "actually" for emphasis | "Tools that actually work" |
| Reference shared frustrations | "Mondays are hard" |
| Keep paragraphs short | 2-3 sentences max |
| Use second person | "Your trial" "helps you" |

#### DON'T ❌

| Pattern | Why Not |
|---------|---------|
| "Leverage" "optimize" "synergy" | Corporate jargon |
| "We are pleased to announce" | Stiff, formal |
| Passive voice | "Was built by our team" |
| Hedging language | "We think maybe possibly..." |
| Long, complex sentences | Loses casual feel |
| Excessive exclamation points | Feels fake enthusiastic |

---

### Word Choice Guide

#### Preferred Terms

| Instead of | Use |
|------------|-----|
| Utilize | Use |
| Leverage | Use |
| Solutions | Tools, software |
| Empower | Help |
| Seamless | Easy, simple |
| Robust | Strong, reliable |
| Best-in-class | (just show, don't tell) |
| Cutting-edge | New, latest |

#### Signature Phrases

| Phrase | When to Use |
|--------|-------------|
| "Here's the thing" | Starting opinion pieces |
| "Actually" | Emphasizing real value |
| "No [negative], no [negative]" | Simple value props |
| "That's backwards" | Calling out industry norms |

---

### Channel Variations

#### Website (Core Voice)
- Most polished version
- Still casual but refined
- Slightly higher stakes language

#### Blog (Extended Voice)
- More conversational
- Can be longer, more exploratory
- Personal opinions encouraged

#### Social Media (Compressed Voice)
- Shortest, punchiest
- More humor allowed
- Emoji occasionally OK
- Can reference current events/culture

#### Email (Direct Voice)
- Friendly but action-oriented
- "Hey there" openings OK
- Clear CTAs
- Personal where possible

#### Support (Empathetic Voice)
- Warmer, more patient
- Solution-focused
- Less attitude, more help
- Still casual but careful

---

### Voice Checklist

Before publishing, check:

- [ ] Would you say this to a friend?
- [ ] Are there any "corporate speak" words?
- [ ] Is it shorter than it could be?
- [ ] Is the main point clear?
- [ ] Does it sound like us, not a competitor?
- [ ] Would it make sense without context?

Example 2: Quick Voice Audit

Input:

Does this match our voice?
"We are delighted to inform you that your request has been
processed and the desired outcome has been achieved."

Output:

## Voice Audit

### Verdict: ❌ Does Not Match

---

### Issues Found

| Problem | Example | Should Be |
|---------|---------|-----------|
| Too formal | "We are delighted" | "Great news!" |
| Passive voice | "has been processed" | "We processed" |
| Corporate speak | "desired outcome" | "what you wanted" |
| Stiff structure | Full sentence | Shorter, direct |

---

### Rewritten

**Original:**
> "We are delighted to inform you that your request has been
> processed and the desired outcome has been achieved."

**On-brand version:**
> "Good news! We've taken care of your request—you're all set."

**Even more casual:**
> "Done! Your request went through. You're good to go."

---

### What Went Wrong

This sounds like a corporate auto-reply, not your brand.
It has:
- No contractions
- Passive voice
- Formal phrasing
- No personality

Your brand would sound like a helpful friend delivering good news,
not a legal notice.

Skill Boundaries

What This Skill Does Well

  • Analyzing existing content
  • Extracting voice patterns
  • Creating guidelines
  • Auditing consistency

What This Skill Cannot Do

  • Create voice from scratch
  • Know your brand strategy
  • Access all your content
  • Replace brand judgment

Iteration Guide

Follow-up Prompts:

  • "Rewrite this in our brand voice"
  • "Create voice variations for [channel]"
  • "Audit these samples for consistency"
  • "Add to our word/phrase dictionary"

References

  • NN/g Voice and Tone Guidelines
  • Content Strategy Alliance
  • Mailchimp Voice and Tone
  • Buffer Brand Voice Guide

Related Skills

  • brand-strategy - Overall brand development
  • copywriting-ogilvy - Writing craft
  • storytelling-storybrand - Narrative voice

Skill Metadata

  • Domain: Branding / Content
  • Complexity: Intermediate
  • Mode: cyborg
  • Time to Value: 2-4 hours for full guidelines
  • Prerequisites: Content samples, brand context

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Measure and optimize growth using the AARRR (Pirate Metrics) framework with stage-specific KPIs and funnel analysis

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

Calculate A/B test statistical significance. Use when: determining if test results are significant; calculating required sample size; estimating test duration; analyzing conversion experiments; making data-driven decisions

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

Assess customer account health using product usage, support sentiment, payment status, and relationship signals

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

Analyze paid advertising performance across channels and recommend budget reallocation to maximize ROAS and minimize CAC. Use when: planning quarterly ad budget allocation, diagnosing underperforming ad channels, deciding whether to scale spend on a channel, calculating marginal ROI across Google Ads, Meta, LinkedIn, or TikTok, rebalancing media mix after performance shifts, or setting up a test-and-scale framework for new channels.

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

Use when testing skills, commands, or agents for quality. Use after creating new skills, before deploying agents, or when debugging inconsistent agent behavior. Triggers on "evaluate", "test quality", "is this skill working", or QA of AI workflows.

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

Use when analyzing server logs to understand how AI crawlers (GPTBot, ClaudeBot, PerplexityBot) interact with your site. Use when optimizing content placement for LLM retrieval, diagnosing why AI search isn't citing your content, or auditing crawl patterns to find optimization gaps.

The English summary is being prepared. Showing the original description.

guia-matthieu/clawfu-skills150Updated Oct 1, 2026

See all skills by guia-matthieu

Report a problem with this skill