voice-check

Calibrate a voice profile and run a contamination linter on drafts.

Updated Apr 14, 2026

Search Tags:#writing #embeddings #calibration #perplexity #discourse-analysis #stylometry #voice-check

Ready to run

npx skills add https://github.com/disabledbydesign/cyborg-methodologies --skill voice-check

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: voice-check
Source: https://github.com/disabledbydesign/cyborg-methodologies/tree/main/voice-check
Command: npx skills add https://github.com/disabledbydesign/cyborg-methodologies --skill voice-check

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Voice-check helps writers preserve their distinctive voice in drafts while automatically surfacing AI-style contamination signals, enabling safer collaboration with AI agents.

Core Features & Use Cases

Calibrates a personalized voice profile from your writing samples
Runs a contamination linter during drafting to surface hedges, jargon, and padding
Learns from revisions to progressively align future drafts with your voice

Quick Start

Run /voice-check setup to create your voice profile from writing samples, then run /voice-check on a draft to verify its voice.

Frequently Asked Questions about voice-check

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

FAQPage Schema

How do I detect AI-style writing in my drafts?▼

To detect AI-style writing in drafts, voice-check runs a contamination linter that surfaces hedges, jargon, and padding. It calibrates a personalized voice profile from your writing samples and returns actionable diagnostics to flag AI-style contamination signals.

What is stylometry and how does it preserve my writing voice?▼

Stylometry analyzes linguistic style to preserve your writing voice by calibrating a personalized voice profile from your samples. Voice-check uses this profile to compare drafts against your unique patterns, ensuring your text reflects your authentic style rather than generic AI outputs.

How do I calibrate a voice profile from my writing samples?▼

You calibrate a voice profile by running the setup command and providing your own writing samples. Voice-check analyzes these texts to establish a baseline of your unique style, which the contamination linter then uses to evaluate future drafts and surface deviations.

Does voice-check work with Claude Code for drafting?▼

Yes, voice-check supports Claude Code integration alongside standalone operation. You can run it within your Claude Code environment to verify drafts during generation, receiving actionable diagnostics and profile updates directly in your workflow.

Can I use discourse-analysis to flag hedges and padding during drafting?▼

Yes, voice-check applies discourse-analysis techniques through a contamination linter to flag hedges, jargon, and padding during drafting. It surfaces these contamination signals in real-time, helping you remove generic AI-style phrasing and align text with your voice.

What are the limitations of using embeddings for voice calibration?▼

When using embeddings for voice calibration, the system learns from your revisions to progressively align future drafts, but it requires sufficient writing samples to establish an accurate baseline. Without enough representative samples, the voice profile may not effectively surface contamination signals.