Developer discovers critical test always passed because it was literally checking 1 = 1

A developer building a production-issue tracking tool realized their most important acceptance test was logically incapable of ever failing. The test compared two counts that were derived from the same underlying filter, making them mathematically identical regardless of whether the tool functioned correctly. When a third-party AI model ran a mutation test — forcing a key function to return null — the test still passed with a perfect score, exposing the flaw. A second rewrite using end-to-end string assertions also failed to catch errors, as the expected output appeared from an unrelated section of the result. The incident highlights how always-green tests can be more dangerous than no tests at all, since they create false confidence and discourage further scrutiny.