Using AI APIs in Your Project

What is a System Prompt?

CONTROL BEHAVIOR · SET PERSONA · SHAPE TONE

🎯 SYSTEM PROMPT BEST PRACTICES

Be specific — "You are a senior JavaScript developer" works better than "You are helpful." Set tone rules  "Use clear examples. Be patient. Explain the 'why' not just the 'how'." Set boundaries  "If you don't know, say you don't know. Never invent APIs."

Build the AI Helper

OPENAI SDK · ASYNC/AWAIT · REUSABLE FUNCTION

📜 CODE BLOCK 1: Complete AI Helper (ai-helper.js with line-by-line comments)

// ============================================
// AI HELPER — Query OpenAI from Node.js
// ============================================

// STEP 1: Install the OpenAI package first
// $ npm install openai

// STEP 2: Import the OpenAI SDK
// WHY: require() loads the openai library
const OpenAI = require("openai");

// STEP 3: Initialize the client with your API key
// WHY: The client handles authentication and API calls
// HOW: Get your key from https://platform.openai.com/api-keys
// PRO TIP: Never hardcode keys! Use environment variables in production
const client = new OpenAI({ 
  apiKey: "YOUR_API_KEY_HERE"  // Replace with your actual key
});

// ============================================
// REUSABLE AI FUNCTION
// ============================================

// WHY: async because we're waiting for the API response
// WHY: Takes a question string, returns AI answer as a string
async function askAI(question) {
  
  // WHY: Call the chat completion endpoint
  // HOW: We specify model, messages array with system and user roles
  const response = await client.chat.completions.create({
    
    // model: gpt-3.5-turbo is fast, affordable, and great for most tasks
    model: "gpt-3.5-turbo",
    
    // messages: array of conversation turns
    messages: [
      {
        // SYSTEM PROMPT — Controls the AI's persona and behavior
        // This turns GPT from a general model into a specialized assistant
        role: "system",
        content: "You are a helpful JavaScript coding assistant for beginner developers. Explain concepts clearly, provide examples, and be patient."
      },
      {
        // USER PROMPT — The actual question from the user
        role: "user",
        content: question
      }
    ]
  });
  
  // WHY: Extract just the answer text from the API response
  // HOW: response.choices[0].message.content contains the AI's reply
  return response.choices[0].message.content;
}

// ============================================
// TEST THE FUNCTION
// ============================================

async function main() {
  // WHY: Ask a real JavaScript question
  // The AI will respond with multiple methods to reverse a string
  const answer = await askAI("How do I reverse a string in JavaScript? Show me 3 methods.");
  
  // WHY: Print the AI's response to the terminal
  console.log(answer);
}

// Run the main function
main().catch(error => console.error("Error:", error));

// ============================================
// HOW TO RUN
// ============================================
// 1. Save as ai-helper.js
// 2. Run: node ai-helper.js
// 3. The AI will answer the question in your terminal
    

📜 CODE BLOCK 2: Production-Ready Version (with .env and error handling)

// PRODUCTION VERSION — Secure and robust
// Install: npm install openai dotenv

// Load environment variables from .env file
require("dotenv").config();

const OpenAI = require("openai");

// Read API key from environment variable (secure!)
const client = new OpenAI({ 
  apiKey: process.env.OPENAI_API_KEY
});

// Check if API key exists
if (!process.env.OPENAI_API_KEY) {
  console.error("Error: OPENAI_API_KEY not found in environment variables");
  console.error("Create a .env file with: OPENAI_API_KEY=your_key_here");
  process.exit(1);
}

// Reusable function with custom system prompt option
async function askAI(question, systemPrompt = "You are a helpful JavaScript coding assistant.") {
  try {
    const response = await client.chat.completions.create({
      model: "gpt-3.5-turbo",
      messages: [
        { role: "system", content: systemPrompt },
        { role: "user", content: question }
      ]
    });
    return response.choices[0].message.content;
  } catch (error) {
    console.error("AI API Error:", error.message);
    return "Sorry, I couldn't process your request. Please check your API key and try again.";
  }
}
    

API Key Security

NEVER HARDCODE · USE .ENV · GITIGNORE

🚫 NEVER hardcode API keys in your source code!

If you commit an API key to GitHub, bots will find it within minutes and use it to make requests on your dime. Always use environment variables. Create a .env file, add it to .gitignore, and access keys via process.env.KEY_NAME. Use the dotenv package to load them.

🤖 HOW AI ACCELERATES THIS TOPIC

Ask AI: "Improve this system prompt to make the AI give shorter, more concise answers." Iterate on your prompts with AI help.

Ask: "Write a system prompt that makes the AI act as a code reviewer who focuses only on security vulnerabilities." Specialize your assistant.

Ask: "What are the rate limits for the OpenAI API and how should I handle them in my code?" Learn production best practices.