blog-style

Analyzes sample blog posts to generate a measurable author voice profile.

2.0k|334|Updated Feb 17, 2026
One-click install
npx skills add https://github.com/AgriciDaniel/claude-blog --skill blog-style
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: blog-style
Source: https://github.com/AgriciDaniel/claude-blog/tree/main/brain/.raw/sources/claude-blog-skill/skills/blog-style
Command: npx skills add https://github.com/AgriciDaniel/claude-blog --skill blog-style

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Writing new blog posts that match an existing author's tone, cadence, and vocabulary is difficult without a concrete style baseline. This Skill learns a quantitative voice profile from 5 to 10 existing posts so future drafts preserve the author's style.

Core Features & Use Cases

  • Voice Profile Generation: Runs a local Python learner over markdown, MDX, or text posts to compute sentence length, burstiness, vocabulary richness, passive voice rate, transition-word rate, and signature phrases.
  • Multiple Output Formats: Produces markdown blocks for VOICE.md, JSON for structured persona creation, and baselines consumable by blog-write and blog-persona.
  • Graceful Degradation: Warns on too few samples, skips missing or unsupported files, and returns zeroed metrics for empty samples instead of crashing.
  • Use Case: A content team onboarding a new writer analyzes ten published articles to build a VOICE.md baseline, then uses it to keep all future drafts consistent with the brand's tone.

Quick Start

Ask the AI to learn your writing style by running the style learner on a folder of your existing blog posts and saving the result as a VOICE.md voice profile.

Frequently Asked Questions about blog-style

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

FAQPage Schema
How do I learn an author's writing style from existing blog posts?▼

Run the style learner script with 5 to 10 representative posts as input, for example python3 scripts/style_learn.py <paths> --format markdown. It aggregates metrics like sentence length, vocabulary richness, and tone into a reusable voice profile.

What metrics does a writing style analysis measure?▼

The analysis measures sentence length mean and variance, type-token vocabulary ratio, passive voice rate, transition-word rate, first-person usage, heading-as-question ratio, paragraph length distribution, and signature 2-gram and 3-gram phrases.

Can I generate a VOICE.md file from blog post samples?▼

Yes, run the learner with --format markdown --output VOICE.md to write a ready-to-use voice profile block. This file can then serve as durable project context for future drafting.

What file formats does the style learner accept?▼

The learner accepts individual markdown files, MDX files, plain text files, or a directory containing posts. Unsupported file types are skipped with a warning included in the profile.

What happens if I provide fewer than 5 sample posts?▼

The learner continues with a warning that the profile may be less stable. Missing paths are skipped, and empty samples return zeroed metrics rather than causing a crash.