0ui-labs/coderabbit-mcp-server

CodeRabbit MCP Server

🇩🇪

Overview

The CodeRabbit MCP Server enables integration with CodeRabbit AI for automated code reviews. It provides a variety of tools and resources to streamline the development process.

Features

🔧 Tools

1. Report Generation: Create on-demand CodeRabbit reports for specific time periods
2. Pull Request Analysis: Analyze specific pull requests for security aspects and other criteria
3. Review Configuration: Set up specific settings for code reviews
4. Interactive Commands: Send commands to CodeRabbit during the review process
5. Health Checks: Check the status of the CodeRabbit environment
6. Custom Reports: Create customized reports with specific filters and templates

📚 Resources

• Sample configurations for .coderabbit.yaml
• Reference for all available CodeRabbit commands
• Examples for AST-Grep rules for code analysis
• Templates for self-hosted environment files

Installation

1. Install dependencies:

   npm install
   

2. Compile TypeScript:

   npm run build
   

3. Start the server:

   npm start
   

Configuration

Environment Variables

Make sure the following environment variables are set:

# CodeRabbit API key (for reports)
export CODERABBIT_API_KEY="cr-xxxxxxxxxxxxx"

# Optional: Custom API Base URL
export CODERABBIT_BASE_URL="https://api.coderabbit.ai/api/v1"

Claude Desktop Integration

Add the server to your Claude Desktop configuration:

macOS/Linux: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "coderabbit": {
      "command": "node",
      "args": ["/Users/philippbriese/Documents/warp/lokale_MCP_Server/code_rabbit/dist/MCPServer.js"],
      "env": {
        "CODERABBIT_API_KEY": "cr-xxxxxxxxxxxxx"
      }
    }
  }
}

Important:

• Use absolute paths!
• Insert your real API key
• Restart Claude Desktop after changes

Usage

Available Tools

1. Report Generation

// Generate a CodeRabbit report
{
  "tool": "generate_report",
  "arguments": {
    "from": "2024-05-01",
    "to": "2024-05-15"
  }
}

2. Pull Request Analysis

// Analyze a pull request
{
  "tool": "analyze_pull_request",
  "arguments": {
    "repository": "owner/repo",
    "pullRequestNumber": 123,
    "reviewInstructions": "Focus on security issues"
  }
}

3. Review Configuration

// Configure review settings
{
  "tool": "configure_review_settings",
  "arguments": {
    "repository": "owner/repo",
    "configuration": {
      "pathInstructions": [
        {
          "path": "**/*.ts",
          "instructions": "Review TypeScript code for type safety"
        }
      ],
      "tools": {
        "astGrep": {
          "essentialRules": true,
          "ruleDirs": ["custom-rules"],
          "packages": ["org/security-rules"]
        }
      }
    }
  }
}

4. Interactive Commands

// Send command to CodeRabbit
{
  "tool": "send_review_command",
  "arguments": {
    "command": "generate docstrings",
    "context": "Focus on public API functions",
    "targetFiles": ["src/api.ts", "src/utils.ts"]
  }
}

5. Health Check

// Check CodeRabbit agent status
{
  "tool": "check_health",
  "arguments": {
    "agentUrl": "http://127.0.0.1:8080"
  }
}

6. Custom Reports

// Create custom report
{
  "tool": "create_custom_report",
  "arguments": {
    "template": "Provide summary of all PR activities and code review discussions",
    "dateRange": {
      "from": "2024-05-01",
      "to": "2024-05-15"
    },
    "filters": {
      "repositories": ["repo1", "repo2"],
      "authors": ["alice", "bob"],
      "includeOnlyMerged": true,
      "excludeBots": true
    }
  }
}

Available Resources

Configuration Templates

• coderabbit://config/sample - Sample .coderabbit.yaml
• coderabbit://env/template - Self-hosted .env template

Documentation

• coderabbit://commands/help - Commands reference
• coderabbit://tools/astgrep - AST-Grep examples

CodeRabbit Integration

Interactive Commands

Use these commands in pull request comments:

# Generate docstrings
@coderabbitai generate docstrings

# Ask for explanation
@coderabbitai Why do all of these functions need docstrings?

# Remember rule
@coderabbitai always remember to enforce camelCase

# Provide context
@coderabbitai do not complain about lack of error handling here, it is handled higher up the execution stack.

Configuration Files

.coderabbit.yaml

reviews:
  path_instructions:
    - path: "**/*.js"
      instructions: |
        Review the JavaScript code against the Google JavaScript style guide
    - path: "tests/**.*"
      instructions: |
        Review unit test code for Mocha best practices

  tools:
    ast-grep:
      essential_rules: true
      rule_dirs: ["custom-rules"]
      packages: ["org/awesome-package"]

code_generation:
  docstrings:
    path_instructions:
      - path: "**/*.ts"
        instructions: |
          End all docstrings with "Auto-generated by CodeRabbit."

Self-Hosted Setup

For self-hosted CodeRabbit instances:

# Pull Docker image
cat coderabbit.json | docker login -u _json_key --password-stdin us-docker.pkg.dev
docker pull <docker-registry>/coderabbit-agent:latest

# Start agent
docker run --env-file .env --publish 127.0.0.1:8080:8080 <docker-registry>/coderabbit-agent:latest

# Health check
curl 127.0.0.1:8080/health

Development

Build

npm run build

Development Mode

npm run dev

Tests

npm test

Troubleshooting

API Key Errors

• Make sure CODERABBIT_API_KEY is set
• Verify that the API key is valid (starts with cr-)

Agent Connectivity

• For self-hosted setups: Make sure the agent is running
• Use check_health tool to test the connection

Configuration Issues

• Use the provided templates as a starting point
• Check YAML syntax in .coderabbit.yaml files

Architecture

The MCP Server acts as a bridge between MCP clients (like Claude Desktop) and the CodeRabbit API. It provides:

• Tools: Interactive functions that can be called by the client
• Resources: Static content like configuration templates and documentation
• Secure Communication: All API calls are handled server-side with proper authentication

Tool Flow

1. Client requests a tool execution
2. Server validates input parameters
3. Server makes authenticated API calls to CodeRabbit
4. Server processes and returns formatted results

Resource Flow

1. Client requests a resource
2. Server returns static content (configs, docs, examples)
3. Content can be used directly or as templates

API Reference

Tool Parameters

All tools accept structured JSON input with validation:

• Dates: ISO format (YYYY-MM-DD)
• Repository names: format "owner/repo"
• File paths: relative to repository root
• Commands: natural language instructions for CodeRabbit

Error Handling

The server provides detailed error messages for:

• Invalid API keys
• Network connectivity issues
• Malformed requests
• CodeRabbit service errors

Links

• CodeRabbit Documentation
• MCP Protocol
• CodeRabbit API Reference

License

MIT