brand-voice enforces your brand writing guidelines automatically as Claude Code writes Markdown. Every time Claude uses Write, Edit, or MultiEdit on a .md or .mdx file, the PostToolUse hook checks it against your brand-guidelines.md and signals Claude to fix violations before saving — no manual review needed.
Features:
• Checks forbidden terms, avoid terms, sentence length, passive voice, and Flesch-Kincaid readability grade
• Smart about code: ignores fenced blocks, indented code, inline code spans, and table cells
• MCP server with analyze_readability and apply_suggestions tools for on-demand checking
• CLI for CI pipelines: check all .md files, changed-only mode, GitHub PR annotations, ratchet baseline enforcement
• /brand-voice-setup skill for guided onboarding — researches your brand from the web or interviews you in 4 questions
• .brand-voice-ignore file and <!-- brand-voice-disable-line --> comment for escape hatches
• Visual identity support: colors, fonts, and logo URLs in the same guidelines file
Works for any brand — configure your own vocabulary, tone, and visual identity in a single brand-guidelines.md file.