Functional QA & Responsive Testing
Crawl reports, performance audits and breakpoint tests turned into fixes, not tickets.
- Role
- Designed and built end to end
- Where
- Remedo · healthcare & health-tech clients
- Tools
- Claude Code · Playwright · Lighthouse · Screaming Frog
functional QA per site
Crawl and Lighthouse reports now go straight in as input.
The problem
Functional QA used to arrive as a report from the QA team: Screaming Frog crawls, responsive issues and broken interactions. Each item then had to be checked and fixed by hand in Webflow.
Responsive issues were the slowest part. Every design had to be looked at across breakpoints, one page at a time, and layout bugs were easy to miss until a client found them.
The approach
The AI now gets the Screaming Frog report, the Lighthouse report and a custom test prompt that exercises buttons, links, sections and forms. It decides what needs fixing and applies the fixes itself.
For layout, Playwright tests every design across 5 breakpoints and is connected to Claude, which picks up the issues it finds and fixes them faster than a manual pass could.
How it runs
- CollectScreaming Frog crawl + Lighthouse report.
- TestCustom prompt exercises buttons, links, sections and forms.
- BreakpointsPlaywright checks every design at 5 widths.
- FixClaude decides what to change and applies it.
Impact
- Every design is tested at 5 breakpoints before launch instead of spot-checked.
- The QA report stopped being a to-do list for a person; it's now an input the system acts on.