← ClaudeAtlas

brand-voice-checkerlisted

Review any draft content against the brand config stored at ~/.claude/brand/brand-config.json. Flags tone mismatches, banned word usage, missing brand terminology, and messaging that contradicts stated differentiators. Produces a line-by-line review with severity ratings and suggested rewrites. Trigger on "check this against brand", "voice check", "brand review", "does this sound like us", "review this for brand voice", "tone check", or when a user pastes content and asks if it matches their brand.
otm-skill-sync/brand-voice-checker · ★ 0 · Data & Documents · score 60
Install: claude install-skill otm-skill-sync/brand-voice-checker
# Brand Voice Checker This skill reviews draft content against your saved brand configuration, identifying where the writing deviates from your established voice, uses banned terminology, or misses opportunities to reinforce brand messaging. ## When to use - User says "check this against brand", "voice check", "brand review", or "does this sound like us" - User pastes content and asks if it matches their brand - User wants a tone/voice audit before publishing - User asks "review this for brand voice" or "tone check" - Any request to verify content alignment with brand guidelines ## Prerequisites This skill requires `~/.claude/brand/brand-config.json` to exist. If it doesn't: 1. Inform the user: "I don't have a brand config on file yet. Want me to walk you through setting one up?" 2. If they agree, hand off to the brand-config skill. 3. If they decline, offer a generic review based on any voice direction they provide inline. ## Instructions ### Step 1: Load brand config Read `~/.claude/brand/brand-config.json` and extract: - Voice attributes (the adjectives that define tone) - Preferred terminology and banned words - Key differentiators and memorable statement - Target audience context ### Step 2: Analyze the content Review the provided content against four dimensions: **A. Voice Attribute Alignment** For each voice attribute (e.g., "confident", "approachable"), assess whether the content consistently reflects it. Flag passages that contradict an attribute. - Examp