Skip to main content

Getting Started with GitHub Copilot

GitHub Copilot, powered by OpenAI, transforms the way developers write code by offering smart, context-aware suggestions. Its effectiveness depends not just on its powerful training but also on how well you guide it! 💡

Quick Tips for Better Prompts​

Start Broad, Then Add Details​

Begin with a general description, then list specific needs.

# Good: Start broad
# Create a user authentication system
# - JWT tokens for sessions
# - Password hashing with bcrypt
# - Email validation

Use Examples​

Show Copilot sample inputs and outputs to make your intent clearer.

// Calculate tax for different regions
// Example: calculateTax(100, "US") -> 108.25 (8.25% tax)
// Example: calculateTax(100, "UK") -> 120 (20% VAT)

Break Down Complex Tasks​

Tackle large problems by asking for smaller parts one at a time.

# ❌ Avoid: "Build a complete e-commerce system"
# ✅ Better: "Create a Product class with price validation"

Be Specific​

Avoid vague terms—clearly refer to functions, code blocks, or libraries.

// ❌ Vague: "Fix this function"
// ✅ Specific: "Fix the validateEmail function to handle edge cases like emails with plus signs"

Essential Setup Tips​

1. Open Relevant Files​

Keep related files open in your IDE. Copilot draws context from all open files to provide more accurate suggestions.

Open Relevant Files

Pro Tip

Use #editor in the chat interface to provide GitHub Copilot with additional context on your currently opened files in VS Code.

2. Add Top-Level Comments​

Begin files with a high-level comment explaining the purpose and goals—Copilot uses this to generate relevant boilerplate and structure.

Top-Level Comments

"""
User Management System
Handles user registration, authentication, and profile management
Uses JWT for session management and bcrypt for password hashing
"""

class UserManager:
    # Copilot will now generate contextually relevant methods

3. Set Imports & References​

Manually include the exact import statements for libraries you're using.

Set Imports & References

// ❌ Without imports - Copilot suggests console.log
import { Logger } from 'log4js';

// ✅ With imports - Copilot suggests Logger methods
const logger = Logger.getLogger();

Writing Copilot-Friendly Code​

Use Meaningful Names​

Just as unclear function names confuse your teammates, they confuse Copilot too!

Use Meaningful Names

# ❌ Unclear
def fetchData():
    pass

# ✅ Clear
def fetch_user_profile_by_email():
    pass

Write Scoped Function Comments​

Add concise comments above functions to describe their behavior.

Write Scoped Function Comments

/**
 * Validates email format and checks domain availability
 * Returns validation status and error message if invalid
 */
function validateEmailAddress(email) {
    // Copilot generates contextually accurate validation logic
}

Provide Sample Code​

Include small code examples or unit tests so Copilot learns your style.

def calculate_discount(price, discount_percent):
    """Calculate discounted price."""
    return price * (1 - discount_percent / 100)

# Example usage:
# calculate_discount(100, 10) -> 90.0
# calculate_discount(50, 25) -> 37.5

Getting Better Results​

Iterate on Solutions​

When asking Copilot Chat for help, you can iterate and improve the solution. Copilot remembers both the generated code and your conversation context.

Example: Creating a Fibonacci function

  1. Start: "Create a function to calculate Fibonacci numbers"
  2. Refine: "Make it non-recursive for better performance"
  3. Polish: "Add input validation and improve variable names"

Use Inline Chat (Ctrl/Cmd + I)​

Use the inline chat feature for quick questions or fixes without leaving the code editor.

Use Inline Chat

Quick Access
  • Ctrl/Cmd + I: Opens inline chat
  • Ask quick questions like "Add error handling to this function"
  • Request fixes: "Make this function more efficient"

Key Takeaways​

  • 🎯 Clear prompts = Better code suggestions
  • 📁 Context matters - Keep relevant files open
  • 💬 Comments help - Explain your intent clearly
  • 🔄 Iterate freely - Refine and improve suggestions
  • 🛠️ Setup properly - Imports, references, and structure matter

Ready to dive deeper? Check out our Best Practices and Agent Mode guides! 🚀