coderz/ai-code-review repository overview

⁠AI-Powered Code Review

This project provides automated code review for GitLab merge requests using AI to analyze code quality.

⁠Features

⁠Code Quality Review

The automated code review analyzes each changed file in merge requests for:

1. Bugs and Logical Errors

   • Undefined variables
   • Memory leaks
   • Logic flow issues
   • Error handling problems

2. Security Vulnerabilities

   • SQL injection risks
   • Command injection vulnerabilities
   • Unsafe file operations
   • Authentication/authorization issues
   • Input validation problems

3. Performance Issues

   • Memory usage concerns
   • CPU intensive operations
   • Inefficient algorithms
   • Resource leaks
   • Caching issues

⁠Setup

1. Configure GitLab CI/CD variables:

   • GITLAB_TOKEN: Your GitLab API token
   • OPENAI_API_KEY: Your OpenAI API key
   • OPENAI_MODEL: The OpenAI model to use (defaults to gpt-4)
   • OPENAI_BASE_URL: (Optional) URL for Ollama API endpoint, e.g., http://localhost:11434/v1
   • PROXY: (Optional) HTTP/HTTPS proxy URL for connecting to OpenAI or Ollama, e.g., http://proxy.example.com:8080

2. Add the .gitlab-ci.yml to your repository

3. Create a merge request to see the automated review in action

⁠Using Ollama

This project supports using Ollama⁠ as an alternative to OpenAI's API:

1. Install and run Ollama on your machine or server
2. Set the OPENAI_BASE_URL environment variable to point to your Ollama instance (e.g., http://localhost:11434/v1)
3. Ensure you have models available in your Ollama installation that are compatible with the OpenAI API format

⁠Using a Proxy

If you need to connect to OpenAI API or Ollama through a proxy:

1. Set the PROXY environment variable to your proxy URL (e.g., http://proxy.example.com:8080 or http://username:[email protected]:8080)
2. The application will automatically route all OpenAI API requests through the specified proxy
3. This works for both OpenAI API and Ollama connections

⁠Usage

The code review runs automatically on all merge requests. It will:

• Analyze changed files for quality issues
• Post detailed review comments on the merge request
• Provide specific suggestions for improvements

⁠Local Development

# Install dependencies
npm install

# Run code review locally
npm run code-review

⁠Scripts

• npm run code-review: Run the code quality review
• npm run analyze: Analyze merge request changes
• npm run validate: Validate merge request size
• npm run debug: Debug GitLab configuration