🐰 CodeBunny

The Blunt AI Code Reviewer

CodeBunny is a high-speed AI agent that acts as a Principal Engineer for your Pull Requests. It skips the fluff and tells you exactly what needs to be fixed to meet professional standards.
Install App • View Demo • Report Bug

🚀 Overview

Most AI tools just summarize code changes. CodeBunny audits them. By deconstructing your code structure before analysis, it focuses strictly on logic, architectural risk, and best practices. It behaves like a senior developer who has seen it all and wants your code to be production-ready—now.

🛠️ Key Features

Feature	Description
👔 Principal Persona	Provides blunt, corrective feedback and one-line code examples.
🏗️ Structural Awareness	Understands function boundaries and logic flow via AST mapping.
⚡ Instant Reviews	Powered by Groq LPUs for near-instant feedback on your PRs.
🚥 Quality Gates	Automatically flags missing docs, complex functions, and naming violations.

📦 Tech Stack

• Brain: Llama 3.3 (70B) via Groq for high-speed inference.
• Analysis: Python AST (Abstract Syntax Tree) for deterministic code mapping.
• Linter: Pylint integration for automated code quality scoring.
• Backend: FastAPI & Uvicorn.
• Infrastructure: GitHub Actions & Render.

📥 How to Use

Add CodeBunny to your repo in 2 minutes:

1️⃣ Install the App

Grant the CodeBunny GitHub App access to your repository.

2️⃣ Add the Workflow

Create .github/workflows/codebunny.yml in your repo:

name: '🐰 CodeBunny Agent'  
on:  
  pull\_request:  
    types: \[opened, reopened, synchronize\]

jobs:  
  review:  
    runs-on: ubuntu-latest  
    permissions: { pull-requests: write }  
    steps:  
      \- name: 🔍 Wellness Check  
        id: health  
        run: |  
          STATUS=$(curl \-s \-o /dev/null \-w "%{http\_code}" "https://codebunny-5o9o.onrender.com/")  
          if \[ "$STATUS" \-eq 200 \]; then echo "status=online" \>\> $GITHUB\_OUTPUT; fi

      \- name: 📡 Dispatch Analysis  
        if: steps.health.outputs.status \== 'online'  
        run: |  
          curl \-X POST "https://codebunny-5o9o.onrender.com/webhook/github" \\  
          \-H "Content-Type: application/json" \-d '${{ toJSON(github) }}' \--fail \--max-time 30

🧪 Local Testing

Test the logic on your machine:

1. Clone & Install

   git clone https://github.com/leonado10000/CodeBunny.git  
   cd CodeBunny  
   pip install \-r requirements.txt

3. Run Audit Simulation

   python local\_run.py

Built with ❤️ by [Rahul Jangra](https://www.google.com/search?q=https://rahul-jangra-leonado10000.vercel.app/projects/codebunny) \