How CodeGuru Reviewer works - Amazon CodeGuru Reviewer

How CodeGuru Reviewer works

Amazon CodeGuru Reviewer uses program analysis combined with machine learning models trained on millions of lines of Java and Python code from the Amazon code base and other sources. When you associate CodeGuru Reviewer with a repository, CodeGuru Reviewer can find and flag code defects and suggest recommendations to improve your code. CodeGuru Reviewer provides actionable recommendations with a low rate of false positives and might improve its ability to analyze code over time based on user feedback.

You can associate CodeGuru Reviewer on a repository to allow CodeGuru Reviewer to provide recommendations. After a repository is associated with CodeGuru Reviewer, CodeGuru Reviewer automatically analyzes pull requests that you make, and you can choose to run repository analyses on the code in your branch to analyze all the code at any time. Recommendations from pull request and repository analysis scans can be viewed directly in the CodeGuru Reviewer console. Recommendations from pull requests can also be viewed as pull request comments in your repository source provider. These recommendations address instances where the code doesn't adhere to AWS SDK best practices, operations on concurrent data structures might not be thread safe, or resource closure might not be handled properly, among other things.

Developers can decide how to incorporate the recommendations from CodeGuru Reviewer and provide feedback to CodeGuru Reviewer about whether the recommendations were useful. This helps your team ensure code quality and improve their code practices in an organic, interactive way. At the same time it improves the quality of recommendations CodeGuru Reviewer will provide for your code, making CodeGuru Reviewer increasingly effective in future analyses.