voice-extractor
Extract voice patterns from transcripts - speaking style, phrases, vocabulary for authentic author profiles
Security Assessment
About voice-extractor
The Voice Extractor skill is designed to analyze transcripts and extract key features that capture an author's authentic voice. It helps transform transcript data from podcasts, interviews, meetings, or videos into actionable writing guidelines that can be used to generate content that closely matches the author's unique speaking style. This skill is particularly useful for enhancing AI-generated content, ensuring that it sounds natural, personalized, and consistent with the author's voice. By extracting specific voice patterns, the tool enables a more genuine, human-like representation of the author in AI-generated writing.
FAQ
What types of transcripts are compatible with the Voice Extractor?
The Voice Extractor supports various transcript formats, including plain text, timestamped, bracketed, WhatsApp-style transcripts, and SRT subtitles.
Can I use audio or video files without transcripts?
Yes, if you don't have a transcript, you can use transcription services like YouTube auto-captions, Otter.ai, Descript, OpenAI Whisper, or Rev.com to convert audio or video into text.
What data is extracted during the analysis process?
The analysis extracts voice characteristics such as sentence structure, communication style, characteristic expressions, vocabulary usage, and speaking style, which can then be used to enhance an author's profile.
How do I integrate the extracted voice data into an author profile?
You can integrate the extracted voice data by using the `/article-writer:author analyze` command to analyze the transcripts and then merge the results with the author's identity information using the `/article-writer:author add` command.
What are the system requirements for running the Voice Extractor?
The Voice Extractor requires a compatible system where the `bun` runtime and the necessary script files are available. It also requires access to transcript data in supported formats.
Install voice-extractor
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
mwguerra/claude-code-plugins