# 🚀 MCP Ollama Code Analyzer

**A self-hosted, privacy-focused alternative to GitHub Copilot.**
Uses **Ollama** (local LLMs) with **MCP servers** in **Docker** for real-time code analysis, refactoring, and optimization.
Requires a **CUDA-capable GPU** for optimal performance.

![GitHub Stars](https://img.shields.io/github/stars/sikienzl/mcp-ollama-code-analyzer?style=social)
![License](https://img.shields.io/badge/license-MIT-blue.svg)
![Docker Pulls](https://img.shields.io/docker/pulls/sikienzl/mcp-ollama-code-analyzer)

---

## 📌 Features
✅ **Local AI-powered code analysis** (no cloud costs!)
✅ **Dockerized MCP server** (easy setup)
✅ **VS Code integration** (Continue/Roo)
✅ **CUDA acceleration** (for optimal performance)
✅ **Supports multiple Ollama models** (e.g., `llama3`, `codellama`)
✅ **Real-time code improvements** (refactoring, bug fixes, optimizations)

---

## 🛠️ Requirements
- **Docker** & **Docker Compose**
- **NVIDIA Container Toolkit** (for CUDA acceleration)
- **Ollama** (locally installed)
- **CUDA-capable GPU** (e.g., NVIDIA RTX 30xx/40xx, A100, etc.)
- **VS Code** with [Continue](https://continue.dev/) or [Roo](https://github.com/RooCodeInc/roo)

---

## 🚀 Quick Start
### 1. Clone the repository
```bash
git clone https://github.com/sikienzl/mcp-ollama-code-analyzer.git
cd mcp-ollama-code-analyzer
```