Roverly docs
Roverly explores your app, records what it finds, and tells you what changed since last time.
What Roverly does
An agent drives your real app — Android, iOS or web — on a machine you control. It maps the screens, then re-walks them on later builds and reports the differences.
Runs execute on your own hardware via the desktop app, not on Roverly’s servers. That is what lets a run reach a physical device, a simulator, or a browser behind your VPN.
Run types
| Type | What it does | Needs |
|---|---|---|
learn | Maps the app and builds the screen inventory. | Nothing. Start here. |
regression | Diffs the whole app against a pinned production run. Catches visual, navigation and structural changes, and crashes. | A pinned production run. |
feature | Goal-scoped run over one area. Catches crashes and logged errors there. No baseline, so it does not report visual change. | A goal. |
Requirements
- The desktop app, installed and paired. Runs need it online.
- A connected device, simulator or browser target.
- For
regression: alearnrun marked as the production run.