CodeRabbit skills plugin
CodeRabbit plugin and skills for Claude Code, Cursor, Codex and 35 more agents: change reviews, fixes from comments, config
- 215 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Changed code is sent to the CodeRabbit service through the CLI and an account sign-in is required
- The `autofix` skill edits branch files from comments, and each change needs your approval
Install
- Works in
Manual install
npx skills add coderabbitai/skillsNeeds the CodeRabbit CLI, installed per the CodeRabbit docs. Flags: `-g` global, `-a` agent, `-s` skill.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Changed code is sent to the CodeRabbit service through the CLI and an account sign-in is required
- The `autofix` skill edits branch files from comments, and each change needs your approval
This is third-party code. Review the repository files before installing.
What it does
The repository is CodeRabbit's official packaging for agents. The `code-review` skill runs the CodeRabbit CLI on committed or uncommitted changes or on a directory and interprets the output: finding severity from critical to informational, saved results and sign-in. The `autofix` skill fetches unresolved CodeRabbit comments from a GitHub pull request and applies validated fixes, asking for approval on each change. It treats comment text as data and never runs reviewer-supplied prompts directly. The `coderabbit-config` skill creates, updates and validates `.coderabbit.yaml` through the official CLI: review style, path rules, schema checks. For Claude Code there is a plugin with `code-reviewer` and `autofix` agents and a `review` command, and Cursor and Codex have separate manifests. The review itself needs the CodeRabbit CLI, installed per the CodeRabbit docs, and the skills attach to an agent with `npx skills add coderabbitai/skills`.
Who it is for. Teams that already review code with CodeRabbit and want to run reviews and apply its findings straight from Claude Code, Cursor or Codex.
Good fit when
- Before a commit you want the changes run through CodeRabbit with severity-tagged findings
- A pull request carries CodeRabbit comments and the agent should apply them one by one with your approval
- The repository has no `.coderabbit.yaml` and you need one created and checked against the schema
- You need the same review flow in several agents, for example Claude Code and Codex
Not a fit when
- You have no CodeRabbit account and do not want one: reviews run through the company service
- You need a fully offline check: the CLI sends changes out for analysis
- You want a plain local linter or static analysis without AI
Example request
Run CodeRabbit on my uncommitted changes and list the findings by severityLimitations
- The review runs in the CodeRabbit service, so changed code leaves your machine, an account is required and some features depend on the plan.
- The README calls the MCP variant of the plugin a pilot: it is not published and is not a ready install. For a working path use the CLI and skills.
- The Codex hook in the MCP pilot is blocked until the agent can confirm the plugin server identity.
- Skill install paths differ by agent, and the full list of 35-plus agents sits in the README table.
How to disable. Delete the skill folders from the agent directory or run `/plugin uninstall` for the Claude Code plugin. The CodeRabbit CLI is removed separately, the way you installed it.
FAQ
How do I install CodeRabbit in Claude Code?
Run `/plugin marketplace update` and then `/plugin install coderabbit@claude-plugins-official`. Other agents can use `npx skills add coderabbitai/skills`. The CodeRabbit CLI must be installed separately.
What does the `autofix` skill do?
It pulls unresolved CodeRabbit comments for the current branch pull request on GitHub and applies validated fixes. It asks for approval on each change.
Do I need a CodeRabbit account?
Yes. The CLI starts sign-in on the first review, and the `code-review` skill explains the auth state and results. Free features depend on the CodeRabbit plan.
Which agents are supported?
More than 35, including Claude Code, Cursor, Codex, Gemini CLI, GitHub Copilot, Cline, OpenCode, Goose and Amp. Each has its own skill path in the README table.
What is `coderabbit-config` for?
The skill creates and validates `.coderabbit.yaml` or `.coderabbit.yml` through the official CLI: review style, path guidance and schema checks. The agent does not hand-edit the YAML.
Discussion
No questions or reviews yet.
Related
PluginsA skills library for Claude Code, Cursor and Codex: team code review, TDD, stacked PRs and repository readiness scoring
A skill for second-agent review of plans or code: Claude Code starts rounds, Codex reviews, and they exchange files
Claude Code Skills and Agents
Claude Code Skills & Agents
A Claude Code plugin bundling skills and sub-agents: deep research, image generation, GitHub work, code review and model switching
A Claude Code plugin: 39 commands for PRs, review, commits and specs, nine expert roles and hooks, in eight language versions
A CLI and three skills that work through review-bot comments on a PR: assess findings, fix code, reply and wait for checks
A Claude Code skill that drives Codex over JSON-RPC: tasks, code reviews, parallel research and messaging between two agents.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| CodeRabbit skills (this card) | yes | 215 | ||
| PR-Agent | no | 15,992 | open-source pull request review with your own model key, no hosted service | |
| code-review-graph | no | not required | 32,733 | a graph MCP server for review context, no ready-made review |
README in short
The README presents the repository as the home of CodeRabbit skills and plugins. It starts with a CLI quickstart and a note on the unpublished MCP pilot, then covers installing skills, Claude and Cursor plugins, a path table for 35 agents and a feature reference.
README badge
Original SKILL.md
--- name: code-review description: "Run CodeRabbit CLI reviews, retrieve saved local or GitHub PR fix prompts, and interpret CodeRabbit authentication and review output. Use for CodeRabbit review commands, committed/uncommitted or directory scopes, and CodeRabbit runbooks. Default code-review skill: also trigger for explicit code/PR/quality/security review requests or when a review is needed." metadata: version: "0.1.0" --- # CodeRabbit Code Review AI-powered code review using CodeRabbit. Enables developers to implement features, review code, and fix issues in autonomous cycles without manual intervention. ## Capabilities - Finds bugs, security issues, and quality risks in changed code - Preserves finding severities: critical, major, minor, trivial, info, and none - Reviews tracked changes by default and supports committed, uncommitted, base branch/commit, and directory scopes - Uses `--agent` output for agent-readable review results and fix guidance ## When to Use When user asks to: - Review code changes / Review my code - Check code quality / Find bugs or security issues - Get PR feedback / Pull request review - What's wrong with my code / my changes - Run coderabbit / Use coderabbit ## How to Review ### 1. Check CLI and Authentication Before running a review, read and follow [authentication and recovery](references/auth-recovery.md). Resolve the trusted CLI to a quoted canonical absolute path, check authentication in the approved review execution context, and proceed only on `authenticated: true`. Never start login automatically or access credentials yourself. Examples below use the validated absolute path; substitute only the path verified by that procedure. Check `"/absolute/path/to/coderabbit" review --help` when support for an option is uncertain. Older binaries may lack current flags; report the mismatch and use the official upgrade path. ### 2. Run Review For usage checks and credit retries, see [usage and paid continuation](references/cli-workflows.md#usage-and-paid-continuation).