🤖 ai-reviewer

AI-powered code reviewer with OWASP Top 10 checks. Fast. Local. Configurable.

• ⚡ Works offline (Fast Mode) or with any AI provider (Cloud Mode)
• 🔒 OWASP Top 10 security scanning
• 🚀 Parallel processing
• 📊 HTML / SARIF / JSON reports
• 🎯 8 languages supported

---

🚀 Installation

pip install ai-reviewer-cli

Or from source:

git clone https://github.com/briej/ai-reviewer.git
cd ai-reviewer
pip install -e .

---

🤖 AI Mode

Use AI-powered analysis with Ollama or cloud providers:

# Ollama (local, free)
ai-review ./project --mode ai --provider ollama --model llama3.1

# DeepSeek (cloud, 1M free tokens)
ai-review ./project --mode cloud --provider deepseek --api-key sk-xxx

# OpenRouter (Anthropic Claude)
ai-review ./project --mode cloud --provider openrouter --api-key sk-xxx

# Groq (fast inference)
ai-review ./project --mode cloud --provider groq --api-key sk-xxx

Supported AI Providers:

• Ollama - Local, unlimited, free (install separately)
• DeepSeek - 1M free tokens/month
• OpenRouter - Rate limited free tier
• Groq - Rate limited free tier
• Kimi - Trial available
• Qwen - Trial available

---

🐳 Docker

Run without installing anything:

# Build image
docker build -t ai-reviewer .

# Run analysis on current directory
docker run -v $(pwd):/code ai-reviewer /code --mode fast

# AI mode with Ollama (requires Ollama running on host)
docker run -v $(pwd):/code --add-host host.docker.internal:host-gateway ai-reviewer /code --mode ai --provider ollama --model llama3.1

# Generate HTML report
docker run -v $(pwd):/code ai-reviewer /code --mode fast --format html --output /code/report.html

# With custom config
docker run -v $(pwd):/code -v $(pwd)/.ai-reviewer.yaml:/app/.ai-reviewer.yaml ai-reviewer /code

---

🎯 Quick Start

# Fast mode — instant analysis, no AI needed
ai-review ./my-project

# AI mode — Ollama-powered analysis (local)
ai-review ./my-project --mode ai --provider ollama --model llama3.1

# Cloud mode — AI-powered analysis (DeepSeek)
ai-review ./my-project --mode cloud --provider deepseek --api-key sk-xxx

# HTML report
ai-review ./my-project --format html --output report.html

# Parallel processing (8 threads)
ai-review ./my-project --threads 8

---

🔥 Features

Security (OWASP Top 10)

• SQL Injection detection
• XSS vulnerability scanning
• Hardcoded secrets detection
• Weak cryptography checks
• CSRF protection validation

AI Providers

Provider	Free Tier	Setup
Ollama	✅ Unlimited	Local install
DeepSeek	✅ 1M tokens	API key
OpenRouter	✅ Rate limited	API key
Kimi	✅ Trial	API key
Qwen	✅ Trial	API key
Groq	✅ Rate limited	API key

Output Formats

• CLI — Rich terminal output with tables
• JSON — Machine-readable
• HTML — Beautiful web report
• SARIF — GitHub Code Scanning compatible

Supported Languages

Python, JavaScript, TypeScript, SQL, Go, Java, Rust, C/C++

---

📋 Usage Examples

Basic

ai-review ./project

Cloud with specific model

ai-review ./project \
  --mode cloud \
  --provider openrouter \
  --model qwen/qwen-2.5-coder-32b \
  --api-key sk-xxx

CI/CD Integration

ai-review . --mode fast --format sarif --output report.sarif

Filter severity

ai-review ./project --severity critical

Ignore patterns

ai-review ./project --ignore __pycache__ --ignore .git --ignore node_modules

---

⚙️ Configuration

Create .ai-reviewer.yaml in your project root:

mode: fast
threads: 4
severity: all

ignore:
  - __pycache__
  - .git
  - node_modules
  - "*.min.js"

cloud:
  provider: deepseek
  model: deepseek-coder
  timeout: 30

rules:
  - id: no-print
    pattern: "print("
    severity: info
    message: "Remove print statements"
    languages: [python]

---

🔄 GitHub Actions

name: Code Review
on: [push, pull_request]

jobs:
  review:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.12'
      - run: pip install ai-reviewer-cli
      - run: ai-review . --mode fast --format sarif --output report.sarif
      - uses: github/codeql-action/upload-sarif@v3
        with:
          sarif_file: report.sarif

AI-Powered CI/CD:

- run: pip install ai-reviewer-cli
- run: ai-review . --mode ai --provider ollama --model llama3.1 --format sarif --output report.sarif

---

🛠 Pre-commit Hook

# .pre-commit-config.yaml
repos:
  - repo: https://github.com/briej/ai-reviewer
    rev: v1.2
    hooks:
      - id: ai-reviewer

---

📊 Example Output

┌──────────────────────────────────────────────────┐
│ 🤖 ai-reviewer — v1.2                            │
│ OWASP Top 10 | Multi-Cloud | Parallel | Rich CLI │
└──────────────────────────────────────────────────┘

✓ Files found: 23

⚠️  CRITICAL (3)
  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  hardcoded-secret   config.py:12   Hardcoded secret
  sql-injection      db.py:45       SQL Injection
  code-injection     utils.py:89    eval() is dangerous

🔶 WARNING (7)
  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  weak-crypto        auth.py:34     Weak hash
  xss                frontend.js:67 innerHTML vulnerable

💡 INFO (12)
  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  debug              main.py:23     console.log

──────────────────────────────────────────────────
┌────────────────┬───────┐
│ Files analyzed │ 23/23 │
│ Time           │ 0.45s │
│ Critical       │ 3     │
│ Warning        │ 7     │
│ Info           │ 12    │
│ Score          │ 4.2/10│
└────────────────┴───────┘

---

🤝 Contributing

See CONTRIBUTING.md

---

📜 License

MIT