How it works
- A pull request is opened or updated and your CI/CD pipelines run.
- CodeRabbit waits for pipelines to finish (up to a configurable timeout).
- CodeRabbit reads the failure logs and identifies the root cause.
- Inline comments with suggested fixes are posted on the relevant lines of code.
Supported platforms
| Platform | How CodeRabbit reads pipeline output |
|---|---|
| GitHub Actions | Via GitHub Checks β configure timeout under reviews.tools.github-checks |
| GitLab CI/CD | Via GitLab pipeline API, including GitLab Advanced Security SAST/DAST findings |
| CircleCI | Via CircleCI pipeline API (requires integration) |
| Azure DevOps Pipelines | Via Azure DevOps Checks API |
github-checks tool configuration.
What failures are analyzed
CI/CD pipeline analysis covers a broad range of failure types:- Build failures β Docker, Node.js, Java, Python, Go module errors
- Test failures β unit, integration, and end-to-end test output
- Security scan findings β SAST and DAST output from tools like GitLab Advanced Security or custom pipeline steps
- Infrastructure-as-code validation β Terraform, Kubernetes, Ansible, and CloudFormation errors
- Linter and code quality output β any linter running as a pipeline step
Configuration
For GitHub Actions, control how long CodeRabbit waits for checks to finish usingreviews.tools.github-checks.timeout_ms.
If your pipelines take longer than 15 minutes, trigger a manual review once they finish:
Whatβs next
GitHub Checks configuration
Configure the timeout and enable/disable behavior for GitHub Actions pipeline analysis.
CircleCI integration
Connect your CircleCI account so CodeRabbit can read pipeline failure logs and post inline fix suggestions.
Pre-Merge Checks
Define custom quality gates that block merges until your standards are met.