AI Code Reviewer

Automated AI-powered code review for your pull requests

AI Code Reviewer is a GitHub Action that automatically reviews your pull requests using AI. It analyzes code changes, finds potential bugs, security issues, and suggests improvements.

Features

• Multiple AI Providers: OpenAI, Anthropic (Claude), Grok, Gemini, Ollama
• Smart Comments: General summary + inline comments on specific lines
• Highly Configurable: Custom prompts, ignore patterns, focus areas
• Multi-language: Review comments in English, Russian, Spanish, and more
• Fail on Critical: Optionally block PRs with critical issues
• Easy Setup: Single workflow file, no complex configuration

Quick Start

1. Create workflow file

Create .github/workflows/ai-review.yml:

name: AI Code Review

on:
  pull_request:
    types: [opened, synchronize]

jobs:
  review:
    runs-on: ubuntu-latest
    permissions:
      contents: read
      pull-requests: write
    
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0
      
      - name: AI Code Review
        uses: username/ai-code-reviewer@v1
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          openai-api-key: ${{ secrets.OPENAI_API_KEY }}

2. Add your API key

Go to Settings → Secrets → Actions and add your API key:

• OPENAI_API_KEY for OpenAI/GPT models
• ANTHROPIC_API_KEY for Claude models
• GROK_API_KEY for Grok models
• GEMINI_API_KEY for Google Gemini models

3. Open a Pull Request

The action will automatically review your code and post comments!

Configuration

Action Inputs

Input	Required	Default	Description
github-token	Yes	—	GitHub token for PR comments
openai-api-key	No*	—	OpenAI API key
anthropic-api-key	No*	—	Anthropic (Claude) API key
grok-api-key	No*	—	Grok (xAI) API key
gemini-api-key	No*	—	Google Gemini API key
ollama-url	No*	—	Ollama server URL
model	No	gpt-4o	AI model to use
fail-on-critical	No	false	Fail action if critical issues found
config-path	No	.github/ai-review-config.yml	Path to config file
language	No	en	Response language

*At least one AI provider is required

Supported Models

Provider	Models
OpenAI	gpt-5.1, gpt-4.1, gpt-4.1-mini, gpt-4.1-nano, gpt-4o
Anthropic	claude-3-5-sonnet-20241022, claude-3-opus-20240229
Grok	grok-2, grok-2-mini
Gemini	gemini-1.5-pro, gemini-1.5-flash
Ollama	Any model: llama3, codellama, mistral, etc.

Custom Configuration

Create .github/ai-review-config.yml for advanced settings:

# Review behavior
language: en              # en, ru, es, de, zh, ja, fr, pt
review_mode: detailed     # quick, detailed, comprehensive
temperature: 0.3          # Lower = more focused

# Focus areas
focus:
  security: true
  performance: true
  best_practices: true
  code_style: true
  documentation: false
  testing: true

# Severity levels to report
severity:
  critical: true
  warning: true
  suggestion: true
  nitpick: false

# File filters
ignore:
  - "*.min.js"
  - "*.lock"
  - "package-lock.json"
  - "**/node_modules/**"
  - "**/dist/**"
  - "**/generated/**"

include_only:            # If set, only review these patterns
  - "src/**"
  - "lib/**"

# Limits
max_files: 20            # Max files per review
max_diff_size: 5000      # Max diff lines to analyze
max_file_size: 1000      # Skip files larger than this

# Custom instructions for the AI
custom_prompt: |
  This is a fintech application.
  Pay special attention to:
  - PCI DSS compliance
  - SQL injection vulnerabilities
  - Proper encryption of sensitive data

Examples

Using Claude

- uses: username/ai-code-reviewer@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    anthropic-api-key: ${{ secrets.ANTHROPIC_API_KEY }}
    model: claude-3-5-sonnet-20241022

Using Gemini

- uses: username/ai-code-reviewer@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    gemini-api-key: ${{ secrets.GEMINI_API_KEY }}
    model: gemini-1.5-pro

Using local Ollama

- uses: username/ai-code-reviewer@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    ollama-url: http://localhost:11434
    model: codellama

Fail on critical issues

- uses: username/ai-code-reviewer@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    openai-api-key: ${{ secrets.OPENAI_API_KEY }}
    fail-on-critical: true

Review in Russian

- uses: username/ai-code-reviewer@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    openai-api-key: ${{ secrets.OPENAI_API_KEY }}
    language: ru

Security

• API keys are passed via GitHub Secrets (encrypted)
• Keys are never logged or exposed
• Only diff content is sent to AI provider
• Minimal permissions required (contents: read, pull-requests: write)

Development

# Clone the repo
git clone https://github.com/username/ai-code-reviewer
cd ai-code-reviewer

# Install dependencies
npm install

# Run tests
npm test

# Build
npm run build

License

MIT License - see LICENSE

Acknowledgments

• OpenAI, Anthropic, Google, xAI for their AI models
• The GitHub Actions team