GitHub Copilot in VS Code: The AI Coding Veteran Explained

What Is GitHub Copilot?

If AI-assisted coding has a household name, it’s GitHub Copilot. Launched in 2021 by GitHub (owned by Microsoft), Copilot was one of the first mainstream AI coding tools – and it’s had years to mature into something genuinely impressive. It’s powered by OpenAI’s models under the hood, deeply integrated with VS Code, and used by millions of developers worldwide.

While the previous posts in this series focused on Claude, it’s worth understanding Copilot properly – especially if you’re using VS Code, where it feels like a natural part of the editor.

Getting Set Up

Step 1: Subscribe to Copilot

Copilot is a paid tool (after a free trial), though there are some free options:

• Free tier – GitHub recently introduced a free Copilot plan with limited completions per month. Enough to evaluate it.
• Copilot Individual – a monthly subscription for solo developers.
• Copilot Business / Enterprise – for teams and organisations, with extra admin controls and privacy features.
• Students and open source maintainers may qualify for free access – check GitHub’s site.

To subscribe, go to github.com/features/copilot and sign up with your GitHub account.

Step 2: Install the VS Code Extension

1. Open VS Code.
2. Go to Extensions (Ctrl+Shift+X / Cmd+Shift+X).
3. Search for GitHub Copilot and install it. You’ll probably also want to install GitHub Copilot Chat – this is the conversational interface.
4. Once installed, click the Copilot icon at the bottom of VS Code and sign in with your GitHub account.

That’s it. Copilot will immediately start working as you code.

How Copilot Works: The Two Modes

1. Inline Completions (the original Copilot)

This is what Copilot is famous for. As you type code, it suggests completions in grey text – sometimes a line, sometimes an entire function. Press Tab to accept, or keep typing to ignore it.

It’s particularly useful for:

• Completing repetitive patterns (Copilot recognises your code style quickly)
• Writing boilerplate based on a comment you’ve typed
• Filling in predictable code like switch cases, object properties, and loop bodies

Try this: type a comment describing what you want, then press Enter and wait. For example:

// function to validate an email address using regex

Copilot will often write the whole function for you.

2. Copilot Chat

The chat interface, accessible from the sidebar or via Ctrl+Shift+I / Cmd+Shift+I, lets you have a conversation with Copilot – similar to Claude.ai but embedded in VS Code. You can ask questions, ask it to explain code, request changes, and more.

You can also open inline chat directly in the editor with Ctrl+I / Cmd+I to ask questions about the code right where you’re working.

Useful VS Code Shortcuts

• Tab – accept the current inline suggestion
• Escape – dismiss a suggestion
• Alt+] / Option+] – cycle to the next suggestion
• Alt+[ / Option+[ – cycle to the previous suggestion
• Ctrl+Enter – open the Copilot suggestions panel to see multiple alternatives
• Ctrl+I / Cmd+I – open inline chat
• Ctrl+Shift+I / Cmd+Shift+I – open Copilot Chat sidebar

Practical Workflow Ideas

Comment-driven development: Write a detailed comment explaining what a function should do, then let Copilot write it. Review and adjust. This keeps your comments accurate and forces you to think through the logic before writing it.

Test generation: Select a function, open inline chat (Ctrl+I), and type “/tests”. Copilot will generate unit tests for your selected code.

Explain unfamiliar code: Select some code you don’t understand, open chat, and type “/explain”. Copilot will walk you through what it does.

Fix errors inline: When you see a red squiggly error in VS Code, you’ll often see a small Copilot lightbulb appear. Click it for AI-suggested fixes.

Rename refactoring: Ask Copilot Chat to suggest better names for a confusingly named variable or function.

Tips for Beginners

• Don’t accept blindly – Copilot’s suggestions can look very plausible but be subtly wrong. Read them.
• Use it for the boring bits – let Copilot handle boilerplate, CRUD, and repetitive patterns so you can focus on the interesting logic.
• Write descriptive variable names and comments – Copilot uses your existing code for context. Good naming and comments lead to better suggestions.
• Try the /fix, /explain, and /tests slash commands in Copilot Chat – these are purpose-built shortcuts for common tasks.

Tips for Advanced Users

• Use .github/copilot-instructions.md in your repository to give Copilot persistent instructions about your codebase – similar to Claude’s CLAUDE.md.
• Copilot works well with multi-file edits via the Copilot Edits feature (formerly called “Workspace edits”) – you can ask it to make changes across multiple files in one go.
• You can switch between AI models in Copilot Chat – GPT-4o, Claude, and Gemini are all available depending on your plan.

Copilot vs Claude Code: A Quick Comparison

Both tools are excellent, but they have different strengths. Copilot’s inline completion is hard to beat for moment-to-moment coding – it feels effortless once you’re used to it. Claude Code (and Claude Code in VS Code) tends to shine on larger, more complex tasks where you want a genuinely conversational back-and-forth about what to build and how.

Many developers use both: Copilot for the continuous inline completions, and Claude Code for bigger feature work or complex debugging sessions.

Wrapping Up

GitHub Copilot is the most mature AI coding tool in the market, and its VS Code integration is excellent. If you’re a VS Code developer, there’s a very good reason it’s become so popular – the inline completion experience, in particular, starts to feel indispensable once you’ve used it for a few weeks.

Final post in the series coming tomorrow: Cursor – the AI-native editor that’s been turning heads across the development community.