Runhuman connects Claude to real human testers who manually QA your web apps and report back structured results. You write test instructions in plain English, a professional tester executes them, and you get findings you can optionally extract as JSON schemas or file directly as GitHub issues. It's built for user flows, visual checks, mobile responsiveness, and exploratory testing, not unit tests or performance benchmarks. The CLI supports templates, device class selection, and tester pool filters on higher plans. Genuinely useful if you're shipping staging builds and want human eyes on signup forms, checkout flows, or layout quirks that Playwright won't catch, though the per-test cost and turnaround time mean you'll still lean on automated tests for the grunt work.
npx skills add https://github.com/volter-ai/runhuman-skills --skill runhuman-testing