voice-calibrator

What problem does it solve?

Analyze and extract voice patterns from writing samples to create or update robust voice skills with measurable targets and authorship-matching validation. Use when calibrating a new voice, refining an existing voice profile, validating content against a voice profile, or comparing two voices. Do NOT use for content generation, general writing, or non-voice editing.

Core Features & Use Cases

• Calibrate voice profiles from 50+ writing samples via deterministic scripts
• Analyze writing samples to extract quantitative voice metrics (sentence length, punctuation, contractions)
• Generate machine-readable profile.json with measurable targets and tolerances
• Generate SKILL.md voice skills with sample-first architecture for authorship matching
• Validate generated content against voice profile metrics
• Show A/B comparisons between default and calibrated output

Quick Start

Collect 50+ writing samples from diverse sources and run the calibration workflow with your samples.