AI Brand Voice Enforcement: How I Built BrandClaw

AI-generated copy is full of "leverage", "synergy", and "seamless" — words that destroy brand voice.

BrandClaw is an OpenClaw skill that fixes this. It audits AI output against brand specs and rewrites violations.

Built-in templates: Stripe (technical, precise), Apple (minimal, emotional), Nike (active voice).

Try online without installing:
👉 https://brand-claw.vercel.app

Install:

bash
git clone https://github.com/xdz199110-lang/BrandClaw.git ~/.openclaw/workspace/skills/brandclaw

How it works:

Hooks into agent:bootstrap and before_agent_reply

Scores output 0-100. Below 70 = flagged and rewritten.

Pure Markdown + JSON, no API calls needed.

What I'd like feedback on:

What brands should I add next?

Is the "Brand Roast" diagnostic useful?

GitHub: https://github.com/xdz199110-lang/BrandClaw