Skip to main content
| CodeRabbit integrates with 50+ third-party linters and security analysis tools to enhance your code reviews. These tools run automatically in secure sandboxed environments, and CodeRabbit uses their findings to inform its review comments.

Default behavior

Almost all tools are enabled by default. CodeRabbit automatically determines which tools are relevant for each project, for example, PHP linters won’t run on a Java project, and Rust linters won’t run on a Python codebase. Tools are only invoked when the repository contains files they understand. This means most projects require no tool configuration at all. The defaults work well out of the box. Configuration is typically only needed when you want to:
  • Disable a specific tool that is not relevant or causes too much noise for your project.
  • Point to a non-default config file location, for example if your ESLint config is in a custom path rather than the project root.
For details on how a specific tool behaves, what file types it targets, and what configuration options it supports, see that tool’s individual page in the tool catalog.

Configuration methods

Individual tool settings

Add tools to your repository’s .coderabbit.yaml file:
.coderabbit.yaml
Use reviews.tools.<tool>.enabled to enable or disable individual tools. The optional config_file field points to your existing tool-specific configuration file (for example .eslintrc.js or pyproject.toml), letting you control which rules are active and their severity. See specific tool guides for more detailed instructions.

Configuration files

When a tool-specific configuration file is present in your repository (such as .eslintrc.js or pyproject.toml), CodeRabbit uses it as-is. When no configuration file is found, most tools fall back to a profile-based default configuration so reviews still run without any setup required.

Profiles

CodeRabbit offers two review profiles that control tool strictness:
  • Chill: Focuses on critical issues and reduces noise from minor style violations
  • Assertive: Provides comprehensive feedback including style and best practice suggestions
    .coderabbit.yaml

Tool findings in reviews

CodeRabbit doesn’t post each tool finding as its own comment. It passes the findings to the review model as hints, and the model checks each one against the code in context. The model decides which findings lead to review comments and leaves out false positives. When an inline review comment covers lines that have tool findings, CodeRabbit adds a collapsed 🧰 Tools section to the end of that comment. The section groups the findings by tool and shows each finding’s severity, line range, message, and rule ID:
The 🧰 Tools section lists only the findings on the lines that the comment covers. Comments that CodeRabbit collects in the review body, such as nitpicks, don’t include it. A finding that no inline comment covers doesn’t appear in any review comment. To see those findings too, set reviews.review_details to true. The comment then lists tool findings in the changed code under Review details > Additional context used, grouped by tool and file.
On Bitbucket Cloud and Bitbucket Data Center, review comments don’t include the 🧰 Tools section.

What’s next

Tool catalog

Browse all linters, security analyzers, and CI/CD integrations by category and technology.

Tool configuration reference

See every YAML configuration option for each supported tool.

YAML configuration

Full reference for all .coderabbit.yaml configuration options.