How do I specify the weight of each voice?

You can use percentage-based natural language triggers like '70% technical, 30% friendly' or use CLI flags to define exact weight ratios for each source profile.

Where are the generated hybrid profiles stored?

Profiles are saved as YAML files in your local '.aiwg/voices/' directory by default, or can be saved globally in your user configuration using the --global flag.

Can I blend more than two voices at once?

Yes, you can combine three or more profiles. By default, they will be blended in equal proportions unless you specify explicit weights for each.

What happens to 'avoid' vocabulary lists when blending?

The skill uses intersection logic for avoidance; a term is only excluded from the hybrid profile if it is marked as 'avoid' in all source voices.

Voice Blend

Name: Voice Blend
Author: jmagly

byjmagly

•

コンテンツ管理

Combines multiple voice profiles with weighted interpolation to create customized hybrid communication styles for AI agents.

概要

Voice Blend is a sophisticated utility for the Claude Code ecosystem that allows users to merge different writing styles into a single, cohesive hybrid profile. By applying weighted mixing to dimensions like formality, confidence, and warmth, the skill enables fine-tuned control over an agent's output persona. It automatically handles complex tasks such as vocabulary deduplication, signature phrase prioritization, and structural consistency resolution, making it an essential tool for creating voices that balance technical precision with approachable conversational tones.

主な機能

Weighted profile mixing for precise tonal and stylistic control
Seamless CLI integration for generating project-specific or global YAML profiles
Automatic interpolation of tone dimensions including formality and confidence
67 GitHub stars
Intelligent vocabulary merging with preference and avoidance logic
Structural resolution based on dominant voice weights and sentence patterns

ユースケース

Standardizing documentation style by merging multiple team-specific voice profiles into a unified brand voice
Developing an 'Executive Summary' style that mixes professional brevity with conversational warmth
Creating a 'Technical Teacher' voice by blending technical authority with friendly explanation

概要

主な機能

Weighted profile mixing for precise tonal and stylistic control
Seamless CLI integration for generating project-specific or global YAML profiles
Automatic interpolation of tone dimensions including formality and confidence
67 GitHub stars
Intelligent vocabulary merging with preference and avoidance logic
Structural resolution based on dominant voice weights and sentence patterns

ユースケース

Standardizing documentation style by merging multiple team-specific voice profiles into a unified brand voice
Developing an 'Executive Summary' style that mixes professional brevity with conversational warmth
Creating a 'Technical Teacher' voice by blending technical authority with friendly explanation