Development Environment Setup

• Package manager for macOS

  • Homebrew
  • Ruby version manager rbenv
  • macports

• General

  • Z Shell via Oh My Zsh
  • Modern Terminal
  • WezTerm
  • Ghostty
  • How to use macOS Recovery!
  • The Quickest Way to Show/Hide Hidden Files
  • Prevent Future .DS_Store Files (Optional)
  • Modify text files via terminal
  • Modify-text-files-via-Nano

• **🛠 XCode Setup **

  • xcode setup
  • MacSpaceCleaner

• **🛠 CLI shortcuts Setup on Mac -command line interface - **

Standardized terminal configuration for React Native, PHP, WordPress, AI Development Tools, and IDE Launchers

• CLI Standardized Configuration

• PhpStorm CLI Setup on macOS

• Text Editors

  • Sublime Download

• GitHub - SSH Access

  • SSH Setup for Business GitHub vs Personal Account

• Package Managers for Programming Languages

  • CocoaPods
  • NPM
  • [Composer]PhpStorm CLI Setup on macOS
  • etc...

🤖 AI Tools

    📦 OpenCode

• OpenCode - Open source AI coding agent available as a terminal app, desktop app, and IDE extension.

• Docs: OpenCode Documentation

• Run OpenCode

  opencode

• Initialize the project

  /init

• Open model picker

  /models
  

• Configure default model in opencode.json

  {
    "model": "provider/model-id"
  }

🦙 Ollama -Local LLM-

• ollama Run local AI models on your machine.
• Ollama Models list or models library

    ⚙️ Ollama Configuration (One-Time Setup)

Create Configuration File

mkdir -p ~/.ollama
nano ~/.ollama/config.yaml

Content

api:
  enabled: true
  address: 0.0.0.0:11434

Test Configuration

curl http://localhost:11434/v1/models

If JSON containing your models is returned, the OpenAI-compatible API is enabled successfully.

    ⌨️ Useful Ollama Commands

Check Running Processes

ps aux | grep ollama

Start Server

ollama serve

Download Model

ollama pull <model-name>

Example:

ollama pull qwen3:latest

Run Model

ollama run <model-name>

Example:

ollama run qwen3:latest

List Installed Models

ollama list

Show Model Details

ollama show <model-name>

Create Custom Model

ollama create

Remove Model

ollama rm <model-name>

🖥️ Cursor + Local LLM

Overview

• Cursor needs Ollama running when you use a local model.
• Use either Ollama.app or ollama serve from terminal.
• Your config.yaml is permanent and normally only needs to be configured once.

Verify Ollama is running

• http://localhost:11434/v1/models

• http://localhost:11434/

  Details

    🎯 Configure Cursor (One-Time)

OpenAI Settings

1. Open Cursor → Settings → Models → OpenAI
2. Enable Override OpenAI Base URL
3. Set Base URL:

http://localhost:11434/v1

Original value:

https://api.openai.com/v1

4. Set any non-empty API Key:

ollama

5. Add your local models:

• qwen3:latest
• gpt-oss:20b
• llama3.1:8b
• deepseek-r1:7b
• etc...

This only needs to be done once unless you reinstall Cursor.

🏗️ Architecture Summary (My Setup)

macOS
│
├── Ollama (local runtime)
│     ├── deepseek-r1:7b
│     ├── mistral:7b
│     └── llama3.1:8b
│
├── CLI tools
│     ├── ollama run
│     └── groq (optional)
│
└── Cursor IDE
      ├── Local LLM (Ollama)
      └── Cloud LLM (optional)

✔ Hardware Requirements

RAM	Recommended Models
16 GB	Small models (7B)
32 GB	Medium models (13B)
64 GB+	Large models (33B / 70B quantized)

🏆 Model Ranking (Reasoning + Coding)

1. Qwen3 (latest) — Best balance of speed and intelligence
2. GPT‑OSS 20B — Strongest reasoning, but heavier
3. Llama 3.1 (8B) — Excellent general-purpose model
4. DeepSeek‑R1 (7B) — Strong for logic and math
5. Qwen2.5 (7B) — Good but older generation
6. Mistral 7B — Fastest, but weaker reasoning

 

 

 

Author

Amr Ahmed Elghadban (AmrAngry)

Contributing 🤘

All your feedback and help to improve this project is very welcome. Please create issues for your bugs, ideas & enhancement requests, or better yet, contribute directly by creating a PR. 😎

When reporting an issue, please add a detailed instruction, and if possible a code snippet or test that can be used as a reproducer of your problem. 💥

When creating a pull request, please adhere to the current coding style where possible, and create tests with your code so it keeps providing an awesome test coverage level 💪

Code of Conduct

I’m here to share my knowledge and findings as I work every day to improve our apps/demos for the community.

This is a space where we work together, openly and safely, as kind and considerate human beings.

We grow by giving and receiving positive, constructive feedback.

Let’s keep learning and building, one step at a time.

License

MIT License.

Distributed under MIT License. Copyright 2025 Amr Elghadban