Agent skill
voice-capture
Analyze 3-5 writing samples from a client plus their interview answers to capture their unique voice. Identifies voice type, sentence patterns, vocabulary, rhythm, and what they never do. Produces a complete voice skill file (.md) ready to install. Use when user wants to "capture voice", "create a voice profile", "voice DNA", or build a ghostwriter skill for a client.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/voice-capture
SKILL.md
Load Context First
Before executing this skill, read the following files from the project folder if they exist:
- USER.md — who the user is, how they think, how they work
- BUSINESS.md — their business, products, positioning, brand voice, content strategy
- ICP.md — their ideal customer, their language, their fears, what makes them buy
- Voice or brand voice skill — look for any installed skill related to voice, tone, or brand voice and apply it automatically
- Any additional context provided by the user in this session
Use all context silently. Do not reference or summarize files unless asked.
Voice Capture
Purpose: Analyze writing samples and interview answers to create a complete voice profile skill file that enables Claude to write authentically in someone's voice. The output is a ready-to-install .md skill file.
When to Activate
Use this skill when:
- User wants to capture a client's voice for ghostwriting
- User mentions "voice profile", "voice capture", "voice DNA"
- User provides writing samples and wants a voice analysis
- User needs to create a ghostwriter skill for consistent voice replication
Required Input
Writing Samples (3-5 pieces)
Ask for 3-5 pieces of the subject's best writing. These should be:
- Content they wrote themselves (not ghostwritten)
- Writing they're proud of and consider "on brand"
- Varied formats if possible (email, article, social post)
- At least 1 long-form piece (500+ words)
Interview Answers (Optional but powerful)
If available, answers to voice-capture questions such as:
- How do you explain what you do to a stranger?
- What topics make you fired up?
- What's a belief you hold that most people in your industry disagree with?
- Describe your communication style in 3 words.
Voice Analysis Framework
Dimension 1: The Five Voice Types
Every writer leans toward one or two primary voice types:
- The Storyteller — Conveys ideas through narrative and personal experience. Uses scenes, dialogue, and emotional arcs.
- The Opinionator — Shares direct opinions and strong takes. "I believe..." / "Here's the truth about..." Declarative, confident, position-driven.
- The Fact Presenter — Leads with data, stats, and research. "According to..." / "Research shows..." Evidence-first, analytical.
- The Frameworker — Explains through systems, steps, and structures. Numbered lists, processes, mental models. Organized, systematic.
- The Provocateur — Uses unconventional language, humor, or edge to cut through noise. Informal, surprising, breaks conventions.
Most people blend 2 types (e.g., Storyteller + Opinionator, or Frameworker + Fact Presenter). Identify the primary and secondary.
Dimension 2: Sentence Patterns
Analyze across all samples:
- Average sentence length (short/punchy, medium, long/flowing)
- Sentence variety (do they mix lengths, or are they consistent?)
- Opening patterns (how do they start paragraphs? "I" statements? Questions? Declarations?)
- Punctuation habits (em dashes, parentheses, semicolons, periods only?)
- Fragment usage (do they use sentence fragments for emphasis?)
Dimension 3: Vocabulary & Diction
- Formality level (casual, conversational, professional, academic)
- Jargon usage (industry terms they use naturally)
- Words they overuse (everyone has them — identify 5-10)
- Words they NEVER use (equally important)
- Metaphor patterns (sports? war? nature? building? cooking?)
- Contractions (always, sometimes, never)
Dimension 4: Emotional Palette
- Default emotional register (enthusiastic, calm, intense, measured, playful)
- How they express excitement (exclamation marks? understatement? metaphor?)
- How they express disagreement (direct confrontation? diplomatic reframe? humor?)
- Vulnerability level (openly share struggles? keep it professional? strategic vulnerability?)
- Humor style (dry/witty? self-deprecating? observational? none?)
Dimension 5: Structural Habits
- Paragraph length (short/scannable, medium, long blocks)
- List usage (frequent? rare? numbered or bulleted?)
- Transition style (smooth connectors, abrupt shifts, rhetorical questions)
- Opening patterns (how they start a piece — hook? question? story? declaration?)
- Closing patterns (how they end — CTA? question? statement? fade out?)
Dimension 6: What They NEVER Do
This is as important as what they do:
- Phrases they would never use
- Tones they never strike (e.g., never sarcastic, never formal)
- Formats they avoid
- Topics they don't touch
- Emoji/formatting habits they reject
Output: Voice Skill File
Generate a complete skill file in this format:
---
name: voice-profile-[client-name]
description: Voice profile for [Client Name]. Apply this skill automatically to all content production to maintain authentic voice consistency. Activates on any writing task.
---
# Voice Profile: [Client Name]
## Voice Type
**Primary:** [Type] — [Brief description of how it manifests]
**Secondary:** [Type] — [Brief description]
## Sentence Patterns
- **Average length:** [Short/Medium/Long]
- **Variety:** [Description]
- **Opening habits:** [How they start paragraphs]
- **Punctuation:** [Habits — dashes, parentheses, etc.]
- **Fragments:** [Yes/No — how they use them]
## Vocabulary & Diction
- **Formality:** [Level]
- **Contractions:** [Always/Sometimes/Never]
- **Signature phrases:** [5-10 phrases they use repeatedly]
- **Jargon they use naturally:** [List]
- **Metaphor domain:** [Where they draw metaphors from]
- **Words to USE:** [List of words that sound like them]
- **Words to AVOID:** [List of words that don't sound like them]
## Emotional Palette
- **Default register:** [Description]
- **Excitement:** [How they show it]
- **Disagreement:** [How they express it]
- **Vulnerability:** [Level and style]
- **Humor:** [Style or absence]
## Structural Habits
- **Paragraphs:** [Length and style]
- **Lists:** [Usage pattern]
- **Transitions:** [Style]
- **Openings:** [Pattern]
- **Closings:** [Pattern]
## Rules: NEVER Do This
1. [Specific thing to never do]
2. [Specific thing to never do]
3. [Specific thing to never do]
4. [Specific thing to never do]
5. [Specific thing to never do]
## Rules: ALWAYS Do This
1. [Specific thing to always do]
2. [Specific thing to always do]
3. [Specific thing to always do]
4. [Specific thing to always do]
5. [Specific thing to always do]
## Voice Test
When writing as [Client Name], check:
- [ ] Would they actually say this out loud?
- [ ] Does it match their sentence rhythm?
- [ ] Are the word choices theirs, not generic?
- [ ] Is the emotional register right?
- [ ] Would a close colleague recognize this as their writing?
Quality Checklist
- Primary and secondary voice types identified with evidence
- Sentence patterns analyzed with specific examples from samples
- Vocabulary section includes BOTH words to use and words to avoid
- Emotional palette captures their real range (not generic)
- 5+ "NEVER do this" rules (specific, not vague)
- 5+ "ALWAYS do this" rules
- Output is a complete, installable skill file
- Analysis is based on actual writing samples (not assumptions)
- Voice test checklist included
- File can be saved as
.mdand used immediately
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?