reopt QA Documentation
Understand a product, verify user flows, and check improvements
reopt QA is a desktop QA application for understanding a product's structure and user flows, then checking improvements through tests and execution evidence.
Post Architecture describes reopt QA's approach to quality design after implementation. Start by separating observed behavior, intended behavior, and areas that have not been verified.
Start with one flow
The Getting Started guide walks through creating a project and checking one important flow. Local automation currently uses the directly distributed macOS app. Check supported systems and installer availability on the download page.
Find a guide
| Task | Guide |
|---|---|
| Understand the installed app and its data | Desktop app |
| Inspect discovered pages, connections, and page results | Service map |
| Organize checks by user flow, feature, and persona | Managing UX checks |
| Prepare executable tests and user conditions | Scenarios and personas |
| Investigate failures and review trends | Reports and analytics |
| Repeat selected tests at an interval | Scheduled checks |
| Let an external AI read results and request permitted actions | MCP collaboration |
Interpret evidence correctly
- A page is on the map: it was discovered within the collection scope. This does not verify every state or feature.
- A test passed: its assertions passed under the conditions used for that run. Missing assertions remain unverified.
- AI suggested something: it is a draft or a hypothesis to review against execution evidence and intended behavior.
Website animations and sample metrics illustrate workflows. Your project's collected data and test results are available in the app.
Execution and data
The app includes a local execution engine. Execution records are stored on your computer, while accessing a target service or external AI involves network communication. See Architecture and data flow and the FAQ for connection boundaries and costs.
App workflow
Use Overview → Service map → Verification → Improvements → Changes to understand and check the product. AI chat is a shared panel that can include context from the selected record.
Track evidence and retests in Improvements, then compare runs in Changes.