For a general overview of how configuration works in CodeRabbit, see Configure CodeRabbit. For complete reference of all options, see Configuration reference.

​
Quick start

Most teams can start with these essential configurations, then customize further as needed:

Essential settings

Data retention, learnings, and review behavior

Common customizations

Review length, path filters, and tool configuration

​
How to configure

1

Choose your approach

Configure using .coderabbit.yaml file (recommended) or CodeRabbit UI. YAML configuration takes precedence over UI settings.
2

Start with data and privacy

Decide on data retention, caching, and learnings based on your organization’s policies.
3

Customize review behavior

Adjust review length, strictness, and automatic behavior to match your team’s preferences.
4

Set path-specific rules

Add path filters and instructions for different parts of your codebase.
The default settings work for most repositories. Use this guide when initially setting up CodeRabbit or when you’re ready to customize its behavior after using it for a while. Each section links to specific entries in the configuration reference, which specifies whether to configure via CodeRabbit UI or your .coderabbit.yaml file.

​
Data and privacy settings

Control how much data CodeRabbit retains about your repositories. CodeRabbit temporarily stores information to provide faster, more tailored reviews, but you can opt out if your organization has strict data-retention policies. For more about CodeRabbit data policies, see Data privacy and security.

​
Data retention (organization-wide)

This setting is only available at the organization level, not per-repository.
Your organization has a master data-retention switch. When disabled, no repositories can use knowledge base or caching features, even if configured individually. We recommend leaving this enabled unless you require strict data retention policies. Configure: Data retention

​
Cache configuration

CodeRabbit keeps a temporary cache of repository data for faster reviews. Disable if this conflicts with your data-retention policies. Configure: Disable cache

​
Knowledge base retention

CodeRabbit builds a knowledge base of metadata about your repositories, team history, and review preferences. This enables more tailored reviews over time. Configure: Opt out

​
Learnings

CodeRabbit learns your team’s preferences when you teach it during reviews using plain language. These learnings apply to future reviews. Set the scope of which stored learnings CodeRabbit applies: Configure: Learnings

​
Review behavior

Customize how CodeRabbit generates and delivers code reviews.

​
Review length and content

Collapse detailed sections and disable optional features: - Collapse walkthrough: Hide detailed change summary in expandable section - Sequence diagrams: Disable visual diagrams - Poem: Disable fun poems about PRs

​
Code review strictness

CodeRabbit focuses on significant issues and includes less nitpicky tool output in reviews.
Configure: Profile

​
Tool configuration

CodeRabbit uses dozens of industry-standard tools like linters and security analyzers. See full tool list. By default, CodeRabbit considers all available tools. Disable specific tools or provide tool-specific configuration files: 
reviews:
  tools:
    pmd:
      enabled: true
      config_file: /pmd-config/*.yml
Configure: Tools

​
Path-specific configuration

Give CodeRabbit different instructions for different parts of your codebase. See Path-based instructions for detailed guidance.

​
Path filters

Exclude files CodeRabbit should ignore, or limit consideration to specific files. Useful for repositories with generated files or non-code content. 
path_filters:
  - "!**/*.bin" # Ignore binary files
  - "!**/*.csv" # Ignore CSV files
  - "**/src/**" # Only review src/ directories
Configure: Path filters

​
Path instructions

Provide specific review instructions for different file types or locations: 
path_instructions:
  - path: "src/**/*.{ts,tsx,js}"
    instructions: "Review React.js, TypeScript, JavaScript code for best practices. Check for security vulnerabilities like SQL injection, insecure dependencies, and sensitive data exposure."
Configure: Path instructions

​
Documentation and testing instructions

Direct CodeRabbit to generate inline documentation or unit tests for specific paths: Configure: Docstrings and Unit tests

​
Workflow integration

Configure how CodeRabbit integrates with your team’s development workflow.

​
Automatic review behavior

By default, CodeRabbit automatically reviews new PRs and updates to existing PRs against your default branch. Configure: Automatic incremental review

​
Review restrictions

​
Pull request approval

We recommend requiring at least one human reviewer approval, even if CodeRabbit can approve PRs.
By default, CodeRabbit never approves PRs—human reviewers handle all approvals. Enable the request changes workflow to allow CodeRabbit approval after all its comments are resolved. Useful for teams requiring multiple approvals (e.g., CodeRabbit + one human reviewer). Configure: Request Changes Workflow

​
Issue creation integration

If you’ve integrated CodeRabbit with Jira or Linear, you can ask CodeRabbit to create issues from PR comments. For private repositories (default), you can disable this feature entirely if needed. Configure: Integrations

​
What’s next