Reference & SupportFeedback & Learning
Feedback on Reviews
How to react to AI review comments
Feedback on Reviews
Thumbs Up / Down
React to review comments to help us improve:
- 👍 Thumbs up — This comment was helpful
- 👎 Thumbs down — This comment was wrong or unhelpful
Why Feedback Matters
Your feedback helps us:
- Identify false positives — Comments that flagged non-issues
- Improve prompts — Refine how the agent analyzes code
- Prioritize features — Understand what matters to users
How It Affects Future Reviews
Currently, feedback is collected but not yet used for automatic per-user tuning.
Future plans:
- Reduce similar false positives
- Learn your team's preferences
- Adapt severity thresholds
Providing Good Feedback
When reacting negatively, consider:
| If the comment was... | The issue might be... |
|---|---|
| Factually wrong | Hallucination; model error |
| Not applicable | Context missing; framework not detected |
| Too minor | Severity threshold too low |
| Duplicate | Already covered by another comment |