Bug Description
The walkthrough summary and the labeling engine run the estimated review effort calculation independently and, in this case, produced different scores for the same PR. This caused an incorrect complexity label to be applied.
Steps to Reproduce
- Repository has a
.coderabbit.yaml that maps CodeRabbit effort scores to complexity labels (e.g., score 1 → complexity: xs, score 3 → complexity: m).
- Open a PR with a trivial diff (1 file changed, +1 line, -0 lines).
- CodeRabbit runs its review.
Result:
- The walkthrough summary reports:
🎯 1 (Trivial) | ⏱️ ~2 minutes
- The labeling engine applies:
complexity: m (implying a score of 3 — Medium)
Expected Behavior
Both subsystems should use the same effort estimate. If the walkthrough says score 1 (Trivial), the labeling engine should apply complexity: xs, not complexity: m.
Actual Behavior
The labeling engine applied complexity: m while the walkthrough summary clearly showed 🎯 1 (Trivial). The two subsystems produced conflicting scores for the exact same PR.
Environment
- Repository: pickleheads-com/app (private)
- PR: pickleheads-com/app#3098
- Diff size: 1 file changed, +1 line added, 0 lines deleted (adding
contents: read to a GitHub Actions workflow permissions block)
.coderabbit.yaml labeling config: score 1 → complexity: xs, score 3 → complexity: m
Additional Context
The PR in question was a one-line fix to a GitHub Actions workflow (sync-labels.yml). The incorrect complexity: m label was manually corrected to complexity: xs after the discrepancy was noticed by the repository maintainer.
Reported by @flash716 via CodeRabbit chat on pickleheads-com/app#3098.
Bug Description
The walkthrough summary and the labeling engine run the estimated review effort calculation independently and, in this case, produced different scores for the same PR. This caused an incorrect complexity label to be applied.
Steps to Reproduce
.coderabbit.yamlthat maps CodeRabbit effort scores to complexity labels (e.g., score 1 →complexity: xs, score 3 →complexity: m).Result:
🎯 1 (Trivial) | ⏱️ ~2 minutescomplexity: m(implying a score of 3 — Medium)Expected Behavior
Both subsystems should use the same effort estimate. If the walkthrough says score 1 (Trivial), the labeling engine should apply
complexity: xs, notcomplexity: m.Actual Behavior
The labeling engine applied
complexity: mwhile the walkthrough summary clearly showed🎯 1 (Trivial). The two subsystems produced conflicting scores for the exact same PR.Environment
contents: readto a GitHub Actions workflow permissions block).coderabbit.yamllabeling config: score 1 →complexity: xs, score 3 →complexity: mAdditional Context
The PR in question was a one-line fix to a GitHub Actions workflow (
sync-labels.yml). The incorrectcomplexity: mlabel was manually corrected tocomplexity: xsafter the discrepancy was noticed by the repository maintainer.Reported by @flash716 via CodeRabbit chat on pickleheads-com/app#3098.