You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently we choose to use the https://coderabbit.ai/ that provides a bot to code reviews.
The exists CI products works really bad. CodeRabbit open source action works fine with open router LLM, but it cannot answer questions and re-scan whole PR.
Activity
vitonsky commentedon Mar 4, 2026
Currently we choose to use the https://coderabbit.ai/ that provides a bot to code reviews.
The exists CI products works really bad. CodeRabbit open source action works fine with open router LLM, but it cannot answer questions and re-scan whole PR.
Also i've tried another solution, but it can't handle even responses of latest Sonnet: https://github.com/Nikita-Filonov/ai-review/issues/76
The intermediate conclusion is there are no robust solution for an automated code review.
For example, CodeRabbit have no option to customize prompt, to add project context, to scan code deeply.