How to Use Claude Code for Manual Web App Testing
I built a manual QA workflow where Playwright MCP walks through my checklist, clicks buttons, and marks pass/fail while I follow along. Here's how it works.
After shipping my appeal mailer site, I needed manual QA for things unit tests miss. I had Claude generate a readiness checklist, then paired it with Playwright MCP so the agent drives the browser and I log bugs as they surface.
0:00 Why manual QA still matters
0:30 Generating a readiness checklist with Claude
1:06 Using Whisper Flow to add notes inline
1:26 Handing the checklist to Playwright MCP
1:56 Where the agent gets stuck (auth, passwords)
2:21 Logging bugs and iterating on fixes
2:42 Visual feedback loops, front end and back end
Subscribe if you want more of these AI build experiments.
Skill I mentioned (QA checklist): https://github.com/craigdossantos/cla...
#ClaudeCode #PlaywrightMCP #AITools #BuildInPublic #ManualQA