Azure OpenAI Code Reviewer for Azure DevOps

The Azure OpenAI Code Reviewer is an intelligent extension that brings AI-powered code review to your Azure DevOps Pull Requests. Using OpenAI's GPT models or Azure OpenAI, it provides GitHub Copilot-style code review feedback directly within your development workflow.

🎯 What It Does

This extension automatically reviews code changes in Pull Requests and provides intelligent feedback on:

• Code Quality - Identifies potential bugs, code smells, and improvement opportunities
• Best Practices - Suggests better coding patterns and standards
• Security Issues - Flags potential security vulnerabilities
• Performance - Recommends optimizations where applicable

✨ Key Features

🔄 Incremental Review System

• Smart Tracking - Only reviews files that have changed since the last review
• Cost Efficient - Significantly reduces OpenAI API usage for large PRs
• Clean Comments - Automatically removes outdated reviews when files change
• Preserves Context - Keeps relevant existing comments while updating others

🤖 AI-Powered Intelligence

• Multiple Models - Support for GPT-4, GPT-3.5-turbo, and GPT-3.5-turbo-16k
• Context Aware - Reviews changes in context using git diffs
• Accurate Analysis - Focuses only on added, modified, or deleted lines
• Intelligent Filtering - Automatically excludes binary files and irrelevant changes

🌐 Enterprise Ready

• Cross-Platform - Works on Linux, Windows, and macOS build agents
• Azure OpenAI - Full support for Azure OpenAI service
• Self-Signed Certificates - Enterprise certificate support
• Secure Integration - Uses Azure DevOps security tokens

🚀 How It Works

1. Trigger - Runs automatically when a Pull Request is created or updated
2. Analysis - Compares changes against the target branch using git diff
3. Review - Sends code changes to GPT model for intelligent analysis
4. Feedback - Posts review comments directly to the Pull Request
5. Tracking - Maintains metadata to enable incremental reviews

📋 Setup Requirements

Build Service Permissions

Grant your build service permission to contribute to Pull Requests:

System Token Access

For YAML Pipelines:

steps:
- checkout: self
  persistCredentials: true

For Classic Pipelines: Enable "Allow scripts to access the OAuth token" in Agent job properties:

⚙️ Configuration Options

OpenAI Integration

• API Key - Your OpenAI API key (stored as secret variable)
• Model Selection - Choose from available GPT models
• Custom Instructions - Tailored prompts for code review

Azure OpenAI Integration

• Endpoint URL - Your Azure OpenAI deployment endpoint
• API Key - Azure OpenAI resource key
• Model Deployment - Your specific model deployment name

Example Azure OpenAI endpoint format:

https://{resource-name}.openai.azure.com/openai/deployments/{model-name}/chat/completions?api-version=2024-02-15-preview

🎛️ Task Configuration

- task: AOAICodeReview@3
  displayName: 'AI Code Review'
  inputs:
    api_key: $(OPENAI_API_KEY)              # Required: API key
    model: 'gpt-4'                          # Optional: Model choice
    aoi_endpoint: $(AZURE_OPENAI_ENDPOINT)  # Optional: Azure OpenAI endpoint
    enable_incremental_review: true         # Optional: Enable smart incremental mode
    support_self_signed_certificate: false  # Optional: Enterprise certificate support
  condition: eq(variables['Build.Reason'], 'PullRequest')

💡 Best Practices

Cost Optimization

• Use Incremental Mode - Reduces API calls by up to 80% on subsequent reviews
• Choose Right Model - Use gpt-3.5-turbo for cost efficiency, gpt-4 for quality
• Filter File Types - Extension automatically excludes binary and irrelevant files

Team Adoption

• Start Small - Begin with smaller repositories to test and tune
• Review Settings - Adjust model and prompts based on your code standards
• Monitor Usage - Track API costs and adjust frequency as needed

🔧 Installation

1. Install Extension - Get it from the Azure DevOps Marketplace
2. Configure API Keys - Set up OpenAI or Azure OpenAI credentials
3. Add to Pipeline - Include the task in your build definition
4. Test with PR - Create a Pull Request to see it in action

🌟 Benefits

• Faster Reviews - Instant feedback on code changes
• Consistent Quality - AI applies consistent review standards
• Learning Tool - Helps developers learn best practices
• Time Savings - Reduces manual review time for maintainers
• Early Detection - Catches issues before they reach production

Ready to enhance your code review process with AI? Install the extension and start getting intelligent feedback on your Pull Requests today!