ai-quickref

📘 AI-Cheat – AI Powered Developer Cheat Sheets in Terminal

A Command Line Interface (CLI) tool that generates concise cheat sheets (React, Python, JavaScript, etc.) using Groq LLM API. No more Googling → just run a command and get an instant cheat sheet inside your terminal. 🚀

✨ Features

• 🖥️ Run in terminal with ai-cheat <topic> (e.g., ai-cheat react hooks)
• ⚡ Uses Groq LLM API for instant AI responses
• 🎯 Supports Zero-shot, One-shot, Multi-shot, Dynamic prompting
• 🔑 Configurable via .env (API key, model, temperature)
• 📊 Logs token usage after each response
• 🎨 Adjustable creativity (temperature) via CLI flag
• 📚 Built to demonstrate 8 AI Prompting Concepts

📂 Project Structure

|-- .env                 # Store your Groq API key here
|-- package.json          # Project dependencies & scripts
|-- package-lock.json
|-- bin
|   |-- ai-cheat.js       # CLI entry file (runs the tool)
|-- src
|   |-- config.js         # Loads environment variables, API config
|   |-- generator.js      # Core cheat sheet generator (API call)
|   |-- prompts.js        # Stores system + example prompts
|   |-- utils.js          # Helpers (token count, formatting, args)

🔑 Setup

• Clone repo

  git clone https://github.com/kalviumcommunity/AI-CLI-CheatSheet
  cd AI-CLI-CheatSheet
  

• Install dependencies

  npm install
  

• Add .env file

  GROQ_API_KEY=your_api_key_here
  MODEL=llama-3.1-70b-versatile
  

• Link CLI globally

  npm link
  

• Run it

  ai-cheat react hooks
  

🧪 Usage Examples

1. Generate React Hooks cheat sheet

ai-cheat react hooks

2. Python list comprehensions

ai-cheat python list comprehensions

3. With creativity (temperature)

ai-cheat javascript promises --temp 0.8

4. View tokens used

ai-cheat express middleware
# Output includes "Tokens used: 234"

🧠 8 AI Prompting Concepts Covered

This project demonstrates 8 important prompting concepts for GenAI:

• Project README + Explanation → Documentation & video demo
• System + User Prompt (RTFC) → Role, Task, Format, Constraints framework
• Zero-Shot Prompting → No examples, just direct query
• One-Shot Prompting → With one sample cheat sheet
• Multi-Shot Prompting → With 2–3 sample cheat sheets
• Dynamic Prompting → User topic auto-inserted in prompt
• Tokens & Tokenization → Log tokens used per query
• Temperature → Control creativity of responses

📜 License

MIT License © 2025