How does the extracted data improve AI writing?

The data maps directly to author profile fields like tone formality, signature phrases, and preferred vocabulary, allowing Claude to mimic the specific nuances of the speaker.

Can I analyze specific speakers in a multi-person transcript?

Yes, you can list all detected speakers in a file and then run a specific extraction command for the individual you wish to profile.

What types of transcripts does Voice Extractor support?

It supports a wide range of formats including plain text, timestamped logs, SRT subtitles, and WhatsApp chat exports.

How much data is needed for a high-quality voice profile?

For reliable results, the skill recommends at least 100 speaking turns and 5,000+ words across multiple conversation contexts.

Does this skill convert audio files to text?

No, you must provide a text-based transcript. However, it provides guidance on using tools like OpenAI Whisper or YouTube auto-captions to generate those transcripts first.

Voice Extractor

Name: Voice Extractor
Author: mwguerra

bymwguerra

•

콘텐츠 관리

Analyzes transcripts to extract authentic speaking patterns, phrases, and vocabulary for personalized author profiles.

Voice Extractor is a specialized Claude Code skill designed to bridge the gap between spoken word and written content by analyzing transcripts from podcasts, interviews, and videos. It meticulously extracts an individual's unique linguistic fingerprint—including sentence structure, signature catchphrases, and communication style—to create highly accurate author profiles. By automating the analysis of speaking patterns, this tool enables users to generate AI content that maintains a consistent, natural, and authentic human voice across various written mediums, ensuring that AI-generated text truly resonates as personal and human-centric.

주요 기능

01Quantifiable communication style metrics for tone, enthusiasm, and directness

02Deep analysis of sentence structure, variety, and vocabulary richness

03Seamless integration with JSON-based author profiles for consistent content generation

04Multi-format transcript processing including SRT, WhatsApp, and timestamped text

053 GitHub stars

06Automated speaker identification and targeted linguistic filtering

사용 사례

01Enhancing brand voice guidelines with data-driven conversational patterns from real employees

02Generating ghostwritten articles that mirror the specific vocabulary and rhythm of a subject matter expert

03Creating a digital twin for a thought leader based on their podcast and interview history

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mwguerra/claude-code-plugins voice-extractor

For use in Claude.ai and ChatGPT

Download Skill