voice-guardlisted

# Voice Guard Quality gate for marketing copy. Scores a draft for machine-sounding writing (filler openers, hype vocabulary, hedging, em-dash overuse, reflexive three-item lists, and flat sentence rhythm) and points at the exact phrases to fix. Run it on anything another skill generates before it ships. ## When to Use - Check this copy for AI tone - Voice check this landing page - Does this email sound human? - Score this draft before we ship it - Clean up the hype in this post - Run voice guard on [file] ## Workflow 1. **Collect the draft** -- paste copy or point at a file produced by another skill 2. **Score** -- run `python_runtime.voice.check_voice(text)` 3. **Triage** -- group findings by kind; fix filler and hype first (largest penalties) 4. **Rewrite** -- replace flagged phrases with concrete claims: numbers, named outcomes, specific users 5. **Re-score** -- repeat until the band is `ship` (85+) 6. **Log** -- keep the before/after scores with the asset so reviewers see what changed ## Reference Full methodology lives in [README.md](README.md) in this folder: configuration, scoring formulas, output formats, integration points, and worked examples. Read it before producing scored output.