What problem does it solve? Writing consistent on-brand content is hard when you lack a concrete definition of an author's voice. This Skill learns measurable style signals from 5 to 10 existing blog posts and produces a reusable voice profile for drafting and persona building. ## Core Features & Use Cases - Style Learning: Runs a local Python learner over markdown, MDX, or text posts to compute sentence length, burstiness, vocabulary richness, passive voice rate, transition usage, and AI trigger word frequency. - Voice Profile Output: Generates markdown, JSON, or VOICE.md-ready output capturing signature phrases, tone descriptors, and paragraph patterns. - Use Case: A content team feeds 8 posts from their founder's blog into the learner, saves the markdown output as VOICE.md, and uses it as a drafting baseline so future AI-assisted posts match the founder's cadence and tone. ## Quick Start Ask the assistant to learn your writing style by running the blog style learn command on a folder containing at least five of your existing blog posts.