PR Review Agent API 🤖

Automated GitHub Pull Request Review Agent powered by Lyzr AI

An intelligent backend system that analyzes GitHub pull requests and generates comprehensive, actionable code review comments using multi-agent AI reasoning.

🚀 Features

• 🔗 Real-Time Webhook Integration: Automatically analyzes PRs when opened or updated via GitHub webhooks
• 🤖 Automated PR Analysis: Fetch and analyze GitHub PR diffs automatically
• 🧠 Multi-Agent AI Review: Powered by Lyzr Agent Studio for intelligent code analysis
• 🔍 Comprehensive Issue Detection: Identifies logic errors, security vulnerabilities, performance issues, and best practice violations
• 💬 Structured Feedback: Generates clear, actionable review comments with severity levels
• 🎨 Beautiful Web UI: Modern dashboard for testing and monitoring
• 📊 GitHub Integration: Automatically posts review comments to GitHub PRs
• 📚 Review History: Stores all reviews in a database for future reference
• 🌐 RESTful API: Clean, well-documented API endpoints
• 🔒 Production Ready: Built with FastAPI, includes logging, error handling, CORS support, and security

📋 Prerequisites

• Python 3.9+
• GitHub Personal Access Token with repo scope
• Lyzr Agent Studio account and API key
• SQLite (default) or PostgreSQL (optional)

🛠️ Installation

1. Clone the repository

cd "c:\Users\Nirupam\Desktop\PR Agent Backend"

2. Create virtual environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables

The .env file is already configured with your credentials. Verify it contains:

# GitHub Configuration
GITHUB_TOKEN=your_github_token
GITHUB_WEBHOOK_SECRET=your_webhook_secret

# Lyzr Configuration
LYZR_API_KEY=your_lyzr_api_key
LYZR_USER_ID=your_email@example.com
LYZR_AGENT_ID=your_agent_id

# Application
SECRET_KEY=your_secret_key
DATABASE_URL=sqlite:///./pr_reviews.db
PORT=8000

🏃 Running the Application

Quick Start (Recommended)

Windows:

start.bat

Linux/Mac:

chmod +x start.sh
./start.sh

Manual Start

Development Mode:

source venv/Scripts/activate  # Windows: venv\Scripts\activate
python main.py

Or using uvicorn directly:

uvicorn main:app --reload --host 127.0.0.1 --port 8000

Production Mode:

uvicorn main:app --host 0.0.0.0 --port 8000 --workers 4

Access Points

• 🎨 Web UI Dashboard: http://127.0.0.1:8000
• 📖 API Documentation: http://127.0.0.1:8000/docs
• 🔧 ReDoc: http://127.0.0.1:8000/redoc
• ❤️ Health Check: http://127.0.0.1:8000/api/v1/health

🎨 Web UI Dashboard

A beautiful, modern web interface is available for easy testing and interaction:

Features:

• 📊 Real-time system health monitoring
• 🔍 Analyze PRs with a simple form
• � View review history
• 🎯 Beautiful, responsive design
• 📱 Mobile-friendly interface

Access the Dashboard: Open your browser and navigate to: http://127.0.0.1:8000

�📚 API Documentation

Interactive API Docs

• Web UI Dashboard: http://127.0.0.1:8000 ⭐ NEW!
• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc

Key Endpoints

1. Health Check

GET /api/v1/health

Check the health status of the application and its dependencies.

2. Analyze Pull Request

POST /api/v1/reviews/analyze
Content-Type: application/json

{
  "pr_url": "https://github.com/owner/repo/pull/123",
  "post_to_github": true
}

Analyzes a GitHub PR and optionally posts the review to GitHub.

Response:

{
  "review_id": 1,
  "pr_url": "https://github.com/owner/repo/pull/123",
  "status": "completed",
  "summary": "Overall review summary...",
  "total_issues": 5,
  "issues": [
    {
      "file": "src/main.py",
      "line": 42,
      "category": "Security",
      "severity": "HIGH",
      "issue": "SQL injection vulnerability detected",
      "suggestion": "Use parameterized queries",
      "code_example": "cursor.execute('SELECT * FROM users WHERE id = ?', (user_id,))"
    }
  ],
  "positive_feedback": "Good code structure and documentation",
  "pr_metadata": {...},
  "created_at": "2025-11-15T10:30:00",
  "completed_at": "2025-11-15T10:30:45"
}

3. Analyze by Details

POST /api/v1/reviews/analyze-by-details
Content-Type: application/json

{
  "owner": "username",
  "repo": "repository",
  "pr_number": 123,
  "post_to_github": true
}

Alternative endpoint to analyze PR by owner, repo, and PR number.

4. Get Review

GET /api/v1/reviews/{review_id}

Retrieve a specific review by ID.

5. List Reviews

GET /api/v1/reviews?skip=0&limit=50

List all reviews with pagination.

6. Delete Review

DELETE /api/v1/reviews/{review_id}

Delete a review by ID.

🏗️ Project Structure

PR Agent Backend/
├── app/
│   ├── api/              # API routes
│   │   ├── reviews.py    # Review endpoints
│   │   └── health.py     # Health check endpoints
│   ├── core/             # Core configuration
│   │   ├── config.py     # Settings and environment variables
│   │   └── logger.py     # Logging configuration
│   ├── models/           # Database models
│   │   └── review.py     # PRReview and ReviewComment models
│   ├── schemas/          # Pydantic schemas
│   │   └── review.py     # Request/response schemas
│   ├── services/         # Business logic
│   │   ├── github_service.py   # GitHub API integration
│   │   └── lyzr_service.py     # Lyzr Agent integration
│   └── utils/            # Utilities
│       └── database.py   # Database utilities
├── main.py               # Application entry point
├── requirements.txt      # Python dependencies
├── .env                  # Environment variables (configured)
├── .env.example          # Environment template
└── README.md             # This file

🔧 Configuration

Environment Variables

Variable	Description	Required	Default
GITHUB_TOKEN	GitHub Personal Access Token	Yes	-
GITHUB_WEBHOOK_SECRET	Webhook secret for validation	No	-
LYZR_API_KEY	Lyzr API key	Yes	-
LYZR_USER_ID	Lyzr user email	Yes	-
LYZR_AGENT_ID	Lyzr agent ID	Yes	-
LYZR_BASE_URL	Lyzr API base URL	No	https://agent.api.lyzr.app/v3
SECRET_KEY	Application secret key	Yes	-
DATABASE_URL	Database connection string	No	sqlite:///./pr_reviews.db
PORT	Server port	No	8000
LOG_LEVEL	Logging level	No	INFO
MAX_DIFF_SIZE	Max diff lines to process	No	10000

Lyzr Agent Configuration

Your Lyzr Agent should be configured with:

• Agent Name: PR Code Review Agent
• Agent Role: Senior Code Reviewer and Quality Assurance Specialist
• Agent Goal: Provide comprehensive code review feedback
• Agent Instructions: See Lyzr agent configuration section in setup docs

🔗 GitHub Webhook Setup (Automated PR Reviews)

Set up webhooks to automatically analyze PRs when they're opened or updated!

Quick Setup

1. Install ngrok (for local development):

   # Download from https://ngrok.com/download
   ngrok http 8000

2. Copy ngrok URL (e.g., https://abc123.ngrok.io)

3. Configure GitHub Webhook:

   • Go to your repo → Settings → Webhooks → Add webhook
   • Payload URL: https://abc123.ngrok.io/webhook/github
   • Content type: application/json
   • Secret: Your GITHUB_WEBHOOK_SECRET from .env
   • Events: Select "Pull requests"

4. Test it: Open a PR and watch the magic happen! 🎉

📖 Detailed Guide: See WEBHOOK_SETUP.md for complete instructions.

Test Webhook Functionality

# Test webhook endpoint
python test_webhook.py

🧪 Testing

Test with cURL

curl -X POST "http://localhost:8000/api/v1/reviews/analyze" \
  -H "Content-Type: application/json" \
  -d '{
    "pr_url": "https://github.com/owner/repo/pull/123",
    "post_to_github": false
  }'

Test with Python

import requests

response = requests.post(
    "http://localhost:8000/api/v1/reviews/analyze",
    json={
        "pr_url": "https://github.com/owner/repo/pull/123",
        "post_to_github": False
    }
)

print(response.json())

📊 Database

The application uses SQLite by default. The database file pr_reviews.db is created automatically on first run.

Database Schema

pr_reviews table:

• id: Primary key
• pr_url: GitHub PR URL
• pr_number: PR number
• repository: Repository name
• pr_title: PR title
• pr_author: PR author
• status: Review status (pending/in_progress/completed/failed)
• review_summary: Overall review summary
• total_issues: Number of issues found
• review_data: Complete review JSON
• created_at: Creation timestamp
• updated_at: Update timestamp
• completed_at: Completion timestamp
• error_message: Error message if failed

review_comments table:

• id: Primary key
• review_id: Foreign key to pr_reviews
• file_path: File path
• line_number: Line number
• category: Issue category
• severity: Issue severity
• issue_description: Issue description
• suggestion: Suggested fix
• code_example: Code example
• created_at: Creation timestamp

🚨 Error Handling

The API returns standard HTTP status codes:

• 200 OK: Successful GET request
• 201 Created: Successful POST request
• 204 No Content: Successful DELETE request
• 400 Bad Request: Invalid input
• 404 Not Found: Resource not found
• 422 Unprocessable Entity: Validation error
• 500 Internal Server Error: Server error

Error responses include:

{
  "error": "Error type",
  "detail": "Detailed error message",
  "status_code": 500
}

📝 Logging

Logs are output to stdout. In development mode, logs are human-readable. In production mode, logs are in JSON format for easier parsing.

Log levels: DEBUG, INFO, WARNING, ERROR, CRITICAL

🔒 Security

• GitHub token is never exposed in logs or responses
• All sensitive data is stored in environment variables
• CORS is configured for specific origins
• Database credentials are protected
• Input validation using Pydantic schemas

🛣️ Roadmap

• Add webhook support for real-time PR monitoring
• Implement rate limiting
• Add Redis caching
• Support for multiple LLM providers
• Custom review templates
• Slack/Discord notifications
• Review metrics and analytics
• Support for GitLab and Bitbucket

🤝 Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests
5. Submit a pull request

📄 License

This project is part of the Lyzr AI Backend Engineering Intern Challenge.

👥 Author

Built for the Lyzr AI Backend Engineering Intern Challenge

🙏 Acknowledgments

• FastAPI: Modern Python web framework
• Lyzr AI: Multi-agent orchestration platform
• PyGithub: GitHub API wrapper
• SQLAlchemy: SQL toolkit and ORM

📞 Support

For issues or questions:

1. Check the API documentation at /docs
2. Review the logs for error details
3. Check GitHub token permissions
4. Verify Lyzr agent configuration

---

Built with ❤️ using FastAPI and Lyzr AI