PersonaQA

QA agents that use your app like real people.

Get started View on GitHub


PersonaQA is a Claude Code plugin that gives your app a cast of simulated users. James is on a phone with patchy 3G; Margaret reads every word on an iPad with large text; Priya never touches a mouse. Give them a mission, such as “buy one adult ticket for Saturday’s match”, and each one drives a real browser as that person would. Then you get their journals, the problems they hit with screenshots, and a ranked report.

A persona buying a ticket, then the ranked report

On a demo shop with 15 deliberately planted flaws, one run of the five built-in personas found 13 of them. That run would cost about $1.50 at API rates with Sonnet; on a Pro or Max plan it uses your plan’s allowance instead. See Cost and usage.

How it fits together

You write PersonaQA does You get
A mission: a goal in plain English plus a rule that proves it was reached Each persona drives a real browser with their device, network and habits enforced A journal per persona, issues with screenshots, and a ranked report you can export

Where to go next

I want to Read
Install it and run my first persona Getting started
Describe the journeys to test Missions
Add my own personas or change the built-in ones Personas
Test pages behind a login Logins
Understand and share the results Reading the results
Know what runs cost, on a plan or an API key Cost and usage
Run it on every pull request CI with GitHub Actions
Look up a config setting Configuration reference
Fix something that is not working Troubleshooting

Back to top

MIT licence. Personas are simulated users: confirm findings before acting on them.

This site uses Just the Docs, a documentation theme for Jekyll.