Brand Voice Synthesizer
Every blog post Claude writes, in your voice — not generic AI voice. The verbal counterpart to the Design System Synthesizer.
About this skill
Read 5-10 published posts and extract the brand voice — tone, lexicon, sentence patterns, person used, formality, signature phrases, phrases the site never uses. Persist it so every future content-writing skill produces copy that sounds like the brand, not like generic AI.
Read-only by default. Any write operation creates a duplicate first, so the original page stays intact and you have an undo handle. All execution runs through Respira's MCP server with your existing credentials.
What it does
4 capabilities run in one pass — each one focused on a single failure mode.
Voice signal extraction
Person used (I / we / you), formality, sentence length (avg / median / p90), paragraph length, reading grade, tone descriptors.
Lexicon mapping
Preferred words the site reaches for, signature multi-word phrases used repeatedly, and the AVOIDED lexicon — words conspicuously absent that competitor sites would use.
Punctuation + structural patterns
Em dash usage, exclamation marks, oxford comma, opens-with-question habits, ends-with-CTA habits, bullet list and code block usage patterns.
Real-corpus examples
A 'this is your voice' paragraph quoted verbatim from your own corpus, plus a constructed counter-example showing what your voice is NOT.
What you'll need
Respira for WordPress plugin installed on the target site, with a valid API key.
An AI agent that supports Skills + MCP — Claude Code, Codex, Antigravity, or Cursor.
SKILL.md saved where your agent looks for skills (see install panel above).
Community feedback
What people say
Rate this skill and share what worked (or what needs to improve).
Community rating
—
/ 5
0 ratings
Community Ratings
Showing: Most helpful
One agent. Many skills.
Browse the full catalog or build your own — the bar is low and the standards are real.
