Redirecting to authentication...

Documentation

Everything you need to know to get started with Rabbit Whisperer and supercharge your PR review workflow.

Quick Start

5 minutes to setup

Get up and running with Rabbit Whisperer in just a few steps. Follow this guide to configure your AI assistant and start reviewing PRs.

1

Create Your Account

schedule
~30 seconds

Create your free account to get started with Rabbit Whisperer. Choose a plan that fits your needs.

check_circle
Free tier includes 25 MCP tokens per month
check_circle
No credit card required to start
2

Generate API Key

schedule
~1 minute

Navigate to your dashboard and create a new API key.

  1. Go to API Keys in your dashboard
  2. Click "Generate New Key"
  3. Give it a descriptive name (e.g., "Claude Code - Personal")
  4. Copy the key immediately - you won't see it again!
security
Keep it secure: Treat your API key like a password. Never commit it to version control.
3

Configure Your AI Assistant

schedule
~2 minutes

Add Rabbit Whisperer to your AI assistant's MCP configuration. Choose your assistant:

Generic MCP Configuration

For other MCP-compatible AI assistants, use the following configuration pattern:

{
  "mcpServers": {
    "rabbit-whisperer": {
      "type": "http",
      "url": "https://api.rabbitwhisperer.dev/mcp",
      "headers": {
        "Authorization": "Bearer <your_API_key>"
      }
    }
  }
}

Key points:

  • type: Set to http for HTTP-based MCP servers
  • url: The MCP server endpoint URL
  • headers: Include your API key in the Authorization header with the Bearer scheme

Consult your AI assistant's MCP documentation for configuration file location.

4

Test Your Setup

schedule
~1 minute

Restart your AI assistant and verify everything is working.

  1. Completely restart your AI assistant (not just reload)
  2. Open a new chat or conversation
  3. Try this test command:
terminal
Test Command
Review PR comments in https://github.com/owner/repo/pull/123
celebration
You're all set!

If everything is configured correctly, your AI assistant will fetch and analyze PR comments. Continue to the GitHub Setup section to add your repositories.

Next Steps

key

GitHub Setup

Add your repositories to enable PR reviews

arrow_forward
lightbulb

Learn the Workflow

Understand how to review PRs effectively

arrow_forward

GitHub Setup

Required

Connect your GitHub repositories to enable PR review functionality. Your tokens are encrypted and stored securely in your account.

info
How does GitHub authentication work?

GitHub tokens are stored securely in your Rabbit Whisperer account settings and encrypted in our database. You don't need to add tokens to your MCP configuration.

1

Create a GitHub Personal Access Token

schedule
~2 minutes

Go to https://github.com/settings/tokens

  1. Click "Generate new token"
  2. Select "Generate new token (classic)"
  3. Give it a name (e.g., Rabbit Whisperer)
  4. Select the repo scope (this grants access to repositories)
  5. Click "Generate token" at the bottom
  6. Copy the token immediately - you won't be able to see it again!
SSO Organizations: If your repository is in an organization that uses SSO, you'll need to click "Configure SSO" next to the token and authorize it for your organization.
2

Add Repository to Rabbit Whisperer

schedule
~1 minute

Navigate to Settings → GitHub Repositories in your Rabbit Whisperer dashboard.

Click "Add Repository", paste your repository URL and the GitHub token you just created, then click "Save".

3

Test Connection

schedule
~30 seconds

The connection will be tested automatically. If successful, your repository is ready to use!

You can now start reviewing PRs from this repository using the MCP tools.

security
Your tokens are secure
  • All tokens are encrypted with AES-256-GCM before storage
  • Tokens are never exposed in API responses
  • You can revoke access anytime from your settings
  • Each repository has its own token for granular control

Usage Guide

4 steps

Learn how to use Rabbit Whisperer to systematically address PR review comments. Our AI-powered workflow helps you analyze, fix, and resolve code review feedback efficiently.

1

Start the Review

Tell your AI assistant:

"Review PR comments in https://github.com/owner/repo/pull/123"

The AI will fetch all unresolved CodeRabbit comments and show you a list organized by severity (🔴 Critical, 🟠 Major, 🟡 Minor).

2

Automatic Analysis

The AI automatically starts analyzing the first comment. You'll see a detailed analysis like this:

Issue #1: src/auth/login.ts (Line 42)
🔴 Critical
What CodeRabbit says:

Missing error handling for authentication failure.

My analysis:

Reviewed the code and CodeRabbit is correct. The login function doesn't handle authentication errors, which could lead to unhandled exceptions. We should wrap the auth call in a try-catch block.

Is this valid?

✓ YES - This is a real issue that should be fixed

Proposed fix:

Wrap the authentication call in a try-catch block:

  async function login(credentials) {
+   try {
      const user = await authenticateUser(credentials)
      return user
+   } catch (error) {
+     logger.error('Authentication failed', error)
+     throw new AuthenticationError('Login failed')
+   }
  }

Should I apply this fix?

3

Apply Fix & Iterate

If you approve, the AI applies the fix to your local file.

You can continue chatting with the AI to refine the solution: "Make the error message more user-friendly"

When you're satisfied, the AI asks:

"Should I reply to the comment and resolve it?"
4

Resolve & Continue

When you approve, the AI posts a reply to GitHub with ✅ and marks the comment as resolved.

The AI then automatically moves to the next comment and repeats the process until all comments are addressed.

Advanced Tips

  • Skip comments: Say "Skip this one, move to next" if you want to address a comment later
  • Refine solutions: Ask the AI to modify the fix before applying: "Use a more descriptive variable name"
  • Review multiple PRs: Start a new review anytime with another PR URL
  • Update comments: (Coming soon) Ask AI to update a comment you already posted

Troubleshooting

Common Issues

Common issues and their solutions. Can't find your issue? Contact us at support or join our community for help.

FAQ

Frequently Asked

Frequently asked questions about Rabbit Whisperer. Find quick answers to common questions about pricing, features, and integration.

How much does it cost?

Rabbit Whisperer has five tiers:
- Free: $0/forever
- Solo: $6.9/month
- Team: $49/month
- Business: $199/month
- Enterprise: $499/month

See pricing page for full details.

Do I need a GitHub token?

Yes, you need to add your repositories with GitHub tokens in your account settings. Tokens are securely encrypted and stored, and you can manage multiple repositories independently.

Can I use with private repositories?

Yes! Add your private repository with a GitHub Personal Access Token (with repo scope) in your settings. All tokens are encrypted with AES-256-GCM before storage.

What AI assistants are supported?

Any AI assistant that supports the Model Context Protocol (MCP), including Claude Code, Cursor, Windsurf, Continue.dev, and more.

How is my code kept secure?

Your code never leaves your machine. Rabbit Whisperer only handles metadata (file paths, line numbers, comment text) and tokens. The AI assistant reads your code locally.

Can I self-host?

Self-hosting is not currently supported, but we're considering it for enterprise plans. Contact us if you're interested in self-hosting options.

Does it work with other code review tools besides CodeRabbit?

Currently, Rabbit Whisperer is optimized for CodeRabbit comments. Support for other review tools may be added in the future based on user demand.

Can I customize the review workflow?

The AI assistant naturally adapts to your preferences through conversation. You can ask it to skip comments, refine solutions, or change the order of fixes.

Ready to get started?

Sign up for free and start reviewing PRs with AI assistance.