Your First Copilot-Style Completion

Autocomplete shows suggested code in gray right where your cursor is. This is called ghost text. Press Tab to accept, Esc to dismiss, or just keep typing to ignore. That is the entire user interface.

The magic is in what triggers a good suggestion. The model sees your open file, nearby files, and any comment you just wrote. The more signal you leave, the better the guess.

Try this exact experiment

1. 1Open a new Python file named greetings.py
2. 2Type a single comment describing the function you want
3. 3Press Enter and wait two seconds
4. 4The AI will draft the whole function as ghost text
5. 5Press Tab to accept, then read every line

# Function that takes a name and a time of day (morning, afternoon, evening)
# and returns a friendly greeting string.

def greet(name, time_of_day):
    greetings = {
        "morning": "Good morning",
        "afternoon": "Good afternoon",
        "evening": "Good evening",
    }
    prefix = greetings.get(time_of_day, "Hello")
    return f"{prefix}, {name}!"

What to do the moment it finishes typing

1. 1Read every line out loud in your head
2. 2Ask: does the dictionary cover all three cases I named?
3. 3Ask: what happens if time_of_day is something weird like None?
4. 4Run the function with a real name and see the output

Pro tip: type the function signature first

If you type the function name and its parameters first, the AI has to fit its body to your shape. That gives you more control than letting it invent everything. Builders who write signatures first get noticeably better suggestions.

“Writing the first line of a function is the most valuable line of code you can write.”

The big idea: autocomplete is a conversation through your comments and signatures. Leave good breadcrumbs, and the AI will follow them to useful code.