StudioTeam turns on CodeRabbit slop detection, 9 of 70 PRs flagged and sparks debate
Original title: Our AI reviewer now labels PRs as slop and the team is split
The author's team enabled CodeRabbit's slop detection two weeks ago, which adds a walkthrough note and a "slop" label to PRs that look like low-effort AI output, such as 600 lines for a 20-line ticket, duplicate helpers, and tests that only check mocks. It flagged 9 of 70 PRs so far, 8 of which the author felt deserved it. The 9th was mostly hand-written by a junior developer who saw the label before anyone spoke to him, so the team changed it to show the label only after a human review. Half the team now wants it on every repo; the other half sees public "slop" labels as shaming with extra steps.
The author shares first-hand data from CodeRabbit's slop detection and the team split over label visibility, offering a real-world tradeoff in AI code review.
The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.