An open API service indexing awesome lists of open source software.

https://github.com/nitya/learn-vscode-copilot

Exploring Vibe Coding examples with VS Code features.
https://github.com/nitya/learn-vscode-copilot

Last synced: about 2 months ago
JSON representation

Exploring Vibe Coding examples with VS Code features.

Awesome Lists containing this project

README

        

# Learn: VSCode With Copilot

Tools like GitHub Copilot can speed up inner loops and improve developer productivity by helping us streamline the more foundational tasks (e.g., scaffolding projects, writing tests and documentation) so we reclaim time to focus on more complex or creative aspects of application development.

When I think about using them today, I see two modes of operation:

1. **["Vibe Coding"](https://en.wikipedia.org/wiki/Vibe_coding)** · I see this as an _open-ended_ exploratory approach to coding where I have a problem to solve and **instruct** the AI to generate the solution till they meet my requirements. A key aspect is that I don't inspect or understand every code change - rather, I focus on staying in the zone and prototyping quickly to validate an idea.

1. **["Pair Programming"](https://learn.microsoft.com/visualstudio/ide/ai-assisted-development-visual-studio?view=vs-2022)** · I see this as a _task-oriented_ productivity hack where I have a specific task or desired outcome and instruct AI **on a more granular level** so I can have better control or understanding of generated code and outcomes. A key aspect here is that I bring some domain or codebase specific expertise to assess changes.

As a visual storyteller, I have a simpler analogy. "Vibe coding" is about painting broad strokes quickly on a canvas so you can see the big picture. "Pair programming" is about adding fine brushstrokes and detail with care, to create unique works of art. _Both approaches have value_ based on goals and context.

The main takeaway is that AI-assisted development is here to stay - so it's time to skill up on the tools, technologies, and best practices. In this project, I want to focus on a single environment (Visual Studio Code) and a single AI tool (GitHub Copilot) - and build up my intuition with hands-on exercises. Join me.

## Pre-Requisites

I'll update this as I go. For now, you will need:

- A GitHub account (with access to Github Copilot and Github Codespaces)
- Familiarity with Python and Jupyter notebooks
- Familairity with Visual Studio Code and Dev Containers

## Getting Started

This repository is configured with a `devcontainer.json`.

- Fork the repo to get a personal sandbox version
- Launch GitHub Codespaces on your fork of the repo
- When ready, open a VS Code terminal and type `mkdocs serve`

This should launch a preview server with instructions to follow.

## Learning Resources

These are some of the core resources I am using for my #30Days project.

- [GitHub Copilot Quickstart](https://docs.github.com/en/copilot/quickstart)
- [GitHub Copilot in VS Code](https://code.visualstudio.com/docs/copilot/overview)
- [GitHub Copilot Patterns and Exercises](https://patterns.hattori.dev/)

Some aspects I am interested in exploring more:

- How can we help AI understand our codebase better? e.g., [llms-txt](https://llmstxt.org/)
- How can we streamline AI usage for documentation? e.g., [LinkedIn tips](https://www.linkedin.com/advice/3/what-best-ways-organize-technical-documentation-uwuqc)
- How can we cite AI-generated content for transparency? e.g., [Purdue resources](https://guides.lib.purdue.edu/c.php?g=1371380&p=10135074)

---

> [!IMPORTANT]
> This is a work in progress. Last updated: Mar 30, 2025.