Documentation
How GreatQA works, what it needs and how to run it.
What GreatQA is
GreatQA is your AI QA employee. It tests changes to your web app in a real browser, then reports back with a result in plain words, screenshots, evidence and a short narrated video. It has two parts:
- The
qacommand tests a branch, ticket, pull request or scenario, by hand or automatically for every merge. - Regression suites are automated tests for every area of your app, written by GreatQA and run in your CI.
Both share a library of reusable test steps that grows with every run.
What it needs
- Your repositories, and a staging or test environment of your app.
- An API key from your AI provider: Anthropic, OpenAI or Google.
- Access to your ticket system, docs and team chat, so it can read context and publish results.
- For parallel runs, a machine that runs private copies of your app. We set this up with you.
Founding customers get all of this set up with them. You don't need to install anything to start.
Results
| Result | Meaning |
|---|---|
| Works as described | The change does what the ticket describes, and nothing nearby broke. |
| Found a problem | At least one finding, with expected against actual and where it comes from. |
| Blocked | Only when something outside your code blocked the test, and it is named in the report. |
Next steps
Read the quick start to see a first run, or the command line reference for every command.


