What problem does it solve? When you want content written in a specific person's style, generic brand-voice profiles fall short because they are not built from real source material. This Skill reverse-engineers a person's writing voice from their actual posts, articles, or tweets, producing a structured voice file that other skills can use to write in that voice. ## Core Features & Use Cases - Parallel Multi-Dimension Analysis: Launches 10 Sonnet subagents in a single message, each examining one dimension such as sentence structure, vocabulary, emotional register, authority signals, and anti-patterns. - Structured Voice File Output: Synthesizes agent findings into a Markdown voice guide with sections for sentence architecture, vocabulary lists, teaching style, tweet voice, and adaptation to your brand. - Per-Subject Voice Storage: Saves each extracted voice to brand/voices/[name].md (or ./voices/[name].md standalone) without overwriting the canonical brand/voice-profile.md. - Use Case: Paste five essays from a founder you admire, and receive a complete voice architecture document plus example tweets adapted to your product, ready for use by writing skills. ## Quick Start Paste several posts or articles from the person whose voice you want to replicate and ask the agent to extract their writing voice into a voice file.
