create-voice

Build a writer's voice profile from samples and save it to .claude/voices/<name>.md.

1|Updated Feb 9, 2026
One-click install
npx skills add https://github.com/brennacodes/brenna-plugs --skill create-voice
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-voice
Source: https://github.com/brennacodes/brenna-plugs/tree/main/plugins/mark-my-words/skills/create-voice
Command: npx skills add https://github.com/brennacodes/brenna-plugs --skill create-voice

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Capture a writer's distinctive voice by building a structured, reusable voice profile from writing samples.

Core Features & Use Cases

  • Load configuration from .claude/mark-my-words.local.md to locate settings.
  • Ensure the .claude/voices/ directory exists to store profiles.
  • Determine the voice name from input arguments or prompt the user, using kebab-case.
  • Validate and avoid overwriting existing profiles unless requested.
  • Gather writing samples from files, pasted text, or both; samples feed the six-profile analysis.
  • Analyze samples to distill Tone & Register, Sentence Patterns, Vocabulary, Rhetorical Habits, Structural Preferences, and Things to Avoid, then save to .claude/voices/<name>.md.
  • Optionally set the profile as the default voice for future posts.

Quick Start

Provide a voice name and a set of writing samples to generate and save your voice profile.

Frequently Asked Questions about create-voice

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a voice profile from my writing samples?

To build a voice profile, provide your writing samples via pasted text or files. The system analyzes them to distill tone, sentence patterns, vocabulary, and structural preferences, saving the result as a markdown file for reuse.

What does a voice profile capture from existing content?

A voice profile captures six distinct elements from your content: Tone & Register, Sentence Patterns, Vocabulary, Rhetorical Habits, Structural Preferences, and Things to Avoid. This distillation ensures future writing matches your original style.

Can I use pasted text and files together to create a voice profile?

Yes, you can gather writing samples from both pasted text and file-based sources simultaneously. The system processes all provided samples together to generate a comprehensive and accurate voice profile.

How do I avoid overwriting an existing voice profile?

The system validates existing profiles and avoids overwriting them unless explicitly requested. You will be prompted to confirm or provide a new kebab-case name if a profile with the same name already exists.

Where is my voice profile stored for future content generation?

Your voice profile is stored as a markdown file at .claude/voices/<name>.md. It includes required frontmatter metadata and can be optionally set as the default voice for future posts.

Do I need a style guide to analyze my writing voice?

No, you do not need a pre-existing style guide. The system acts as the style guide generator by directly analyzing your provided writing samples to extract your distinctive rhetorical habits and structural preferences.