An AI-powered code review assistant that runs completely local using Ollama.

Code Minion is a command-line tool that analyzes codebases to find bugs, security issues, and quality problems. It uses local LLMs through Ollama to provide intelligent, context-aware reviews without sending code to external services.

• Fully Local: All analysis runs on your machine using Ollama - no code leaves your system
• Language-Agnostic: Works with any programming language through LLM analysis
• Git-Aware: Analyzes diffs between branches to provide targeted reviews
• Multiple Analysis Types: Combines LLM-based analysis with optional static analyzers
• Clean Reports: Generates markdown reports organized by severity

1. Python 3.12+
2. Ollama installed and running. Prefer to use CodeGemma, deepseekcoder models. They perform better at code analysis and reviews.
3. Git

1. Clone the repository

git clone https://github.com/kanapuli/code-minion.git
cd code-minion

2. Set up a virtual environment using uv

# Install uv if not already installed
pip install uv

# Create and activate virtual environment
uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies

uv install # This will install the dependencies in pyproject.toml

4. Pull a CodeLlama model using Ollama

ollama pull codegemma:7b

# Run from your git repository
python -m code_minion review-diff --base-branch main

# Specify output file
python -m code_minion review-diff --base-branch main --output review.md

# Use a different model
python -m code_minion review-diff --base-branch main --model codellama:13b

# Review specific files
python -m code_minion review-files path/to/file.py another/file.js

# Review all files in a directory
python -m code_minion review-dir src/

Create a .code_minion.json file in your home directory to customize the tool:

{
  "ollama": {
    "base_url": "http://localhost:11434",
    "model": "codellama:7b",
    "temperature": 0.2
  },
  "analysis": {
    "max_files": 100,
    "max_file_size": 1000000,
    "ignore_patterns": [
      "**/.git/**",
      "**/node_modules/**",
      "**/__pycache__/**"
    ]
  },
  "reporting": {
    "format": "markdown",
    "output_file": "code_minion_report.md"
  }
}

Code Minion follows a modular architecture:

• Repository Manager: Handles git operations and file access
• Analyzer Framework: Pluggable system for different analysis methods
• LLM Integration: Structures prompts and parses responses from Ollama
• Formatters: Generates readable reports from analysis results

To create a new analyzer, extend the BaseAnalyzer class:

from code_minion.analyzers.base import BaseAnalyzer
from code_minion.core.models import AnalysisResult

class MyCustomAnalyzer(BaseAnalyzer):
    def __init__(self):
        super().__init__(name="my-custom-analyzer")
        
    def analyze_file(self, file_path, file_content, metadata):
        # Your analysis logic here
        result = AnalysisResult(analyzer_name=self.name)
        # Add issues to result.issues
        return result

MIT License

Contributions are welcome! Please feel free to submit a Pull Request.