Skip to main content
Reaction-based fine-tuning is not currently active. 👍/👎 reactions and suggestion implementation status can be recorded, but they do not train a model, build preference clusters, or filter the next review.
A thumbs-up or thumbs-down alone does not explain why a suggestion was rejected. Richer negative-feedback context is needed before this learning loop can be reintroduced. You can still leave reactions as optional product feedback. Do not expect reacting to a suggestion to automatically suppress similar suggestions in later reviews.

​
What affects future reviews

  • Memories: Persist explicit team conventions through @kody remember and conversations.
  • Kody Rules: Define standards that are applied during code review.
  • Automatic Rules Generation: Propose rules from review history for manual import. This is separate from reaction-based fine-tuning and remains an active feature.