Automated the entire UX copy QA process using n8n & Figma API. I built a workflow that:
- Pulls microcopy directly from Figma
- Filters out junk labels, numbers, all-caps, and fillers
- Runs each string through an LLM for grammar, tone, and clarity review
- Parses structured suggestions from the AI
- Saves all feedback in a Google Sheet (screenshot below)
If you’re part of a product or content team and want to see it in action, I’m happy to give you a demo! 😊
PS. The empty fields are just single text line items, which the LLM is ignoring, only returning the Figma Node IDs. The filter is not perfect yet.