Get the FREE Ultimate OpenClaw Setup Guide →

mcp -workshop

MCP server from Visko93/mcp-server-workshop

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio visko93-mcp-server-workshop npx -y package-name \
  --env MODEL_NAME="deepseek-coder" \
  --env OLLAMA_HOST="http://localhost:11434"

How to use

This MCP server provides a local AI code reviewer that analyzes your code changes using a local model. It leverages Ollama to run the deepseek-coder model and exposes an CLI workflow that you can integrate into your development process. The server is designed to run entirely on your machine without requiring external API keys or internet access, making it suitable for private, offline-style code reviews. To use it, ensure Ollama is installed and that the deepseek-coder model is available. The workflow typically involves invoking the reviewer against a project path; the tool then analyzes uncommitted diffs and prints a structured, colorized set of suggestions and actionable fixes directly in your terminal.

How to install

  • Prerequisites:
    • Windows 11 (or your OS)
    • Node.js v22.0.0 or higher
    • npm v11.0.0 or higher
    • Ollama installed and accessible from your system PATH
  • Installation steps:
    1. Clone the repository:
      git clone <repository-url>
      cd code-reviewer
      
    2. Install dependencies:
      npm install
      
    3. Pull the required model:
      ollama pull deepseek-coder:latest
      
    4. Verify the installation:
      ollama list   # Should show deepseek-coder
      
  • Run the server/tool (example):
    npx tsx src/index.ts /path/to/your/project
    
  • Optional: configure environment or paths as needed for your setup.

Additional notes

Tips and considerations:

  • Local-first operation relies on Ollama and a local model; ensure your hardware meets memory requirements (8 GB+ RAM is often recommended for local ML models).
  • If Ollama or the model doesn’t respond, check that Ollama is running and the model is loaded (ollama list).
  • The reviewer analyzes uncommitted changes in the target project path; ensure you provide the correct path to get accurate diffs.
  • If you encounter Node version issues, verify that you’re using Node.js v22+ and npm v11+ as per the prerequisites.
  • The core output is a colorized terminal report with a Summary and Detailed Analysis; you can redirect output to a file if needed for reviews.

Related MCP Servers

Sponsor this space

Reach thousands of developers