UX Copy Review with AI

# UX Copy Review — SoloStack Skill File
> Drop this file into your SoloStack repo and Claude Code will use it as
> context to run this workflow automatically.
---
## Overview
Refine user-facing copy for clarity, conviction, and conversion. Reviews buttons, headlines, empty states, error messages, onboarding flows, and form labels. Returns specific rewrites with the reason for each change.
## Capabilities
1. **Surface-by-surface rewrites** — Reviews each button, headline, error, empty state, and form label. Returns original vs rewrite side-by-side with reason for each change.
2. **Severity-tagged** — Critical (ship-blocker), important (should fix), nice-to-have (polish). So you know what to prioritise.
3. **Voice alignment** — Reviews against your brand voice guide and the SoloStack AI-tell ban list. Catches em dashes, "leverage", "delve", hype adjectives, hedging.
4. **Anti-pattern detection** — Flags reusable mistakes: jargon dumps, hedged CTAs, apologetic errors, vague verbs. Gives you the replacement pattern.
5. **Reusable patterns extracted** — Lifts rules from the rewrites that apply elsewhere (e.g. "all errors name the fix, not the failure") so the next developer applies them automatically.
6. **Pre-ship checklist** — Short final scan list for the next copy review. Stops the same mistakes shipping twice.
## Pipeline
## Use Cases
- **Onboarding flow polish:** You're about to ship a 12-screen onboarding written by engineers. Review every screen for clarity, voice, and conviction before launch.
- **Conversion drop on pricing page:** Conversion dropped 30% after the last redesign. Review every line of pricing copy and CTAs to find what's killing it.
- **Error message audit:** Support tickets show users confused by error messages. Audit every error and rewrite to name the fix, not just the failure.
- **Brand voice rollout:** Marketing locked a new voice. Product UI still sounds like the old voice. Audit and align.
## Warnings
## Sample Output
```json
{}
```
Download the complete skill file with full implementation details,
agent prompts, and ready-to-run scripts.