Verification Theater
2026.176The deploy broke. The checks were green. The mechanism measures what it can see - and what it can't see is everything that breaks.
code-review
all tags →The deploy broke. The checks were green. The mechanism measures what it can see - and what it can't see is everything that breaks.
Three AI code-review tools sit on the same model. The one that guarantees coverage did it with a for-loop, not a better prompt.
The PR has 47 comments. None of them are about the approach.
The old failure mode was reading every line equally. The new one is trusting code that looks right.
A dev who didn't speak in meetings and whose PRs were the conversation.
When quality looks suspicious, people stop producing quality.
Pattern recognition lives in the practice. The flinch that catches the N+1 query doesn't survive a full calendar.
The bugs are quieter now. Just wrong assumptions buried in correct-looking code.
Phabricator changed what code review means to me. A note from a few months in.