VS Code + GitHub Copilot Tutorial

This tutorial helps you set up and use GitHub Copilot in VS Code to write, refactor, test, and document code faster.

1) Prerequisites

• A GitHub account with Copilot access
• Visual Studio Code (latest recommended)
• Internet connection

2) Install Copilot in VS Code

1. Open VS Code.
2. Go to Extensions (Cmd+Shift+X on macOS / Ctrl+Shift+X on Windows/Linux).
3. Search for and install:
   • GitHub Copilot
   • GitHub Copilot Chat
4. Sign in to GitHub when prompted.

3) Verify Copilot Is Working

• Open any project file (for example, .js, .py, or .ts).
• Start typing a function.
• If Copilot suggests ghost text inline, it is active.
• Press:
  • Tab to accept suggestion
  • Esc to dismiss

4) Core Workflows

A. Inline Code Completion

Use for fast coding while typing:

1. Write a clear function name and comment.
2. Pause briefly for suggestions.
3. Accept full (Tab) or keep typing to refine.

Example comment:

# Parse a CSV file and return rows grouped by country

B. Copilot Chat (In-Editor Assistant)

Use for explanations, refactors, and generation:

1. Open Chat panel (Copilot icon in the sidebar).
2. Ask focused prompts such as:
   • “Explain this function in plain language.”
   • “Refactor this to reduce duplication.”
   • “Write unit tests for this file using pytest.”
3. Review generated code before applying.

C. Edit/Rewrite With Context

1. Select code block.
2. Open Copilot Chat.
3. Prompt with intent + constraints:
   • “Rewrite this to be async, keep the same behavior, and preserve function names.”

5) Prompting Tips (High Impact)

• Be specific about:
  • language/framework
  • desired output format
  • constraints (performance, style, compatibility)
• Provide context files or snippets when possible.
• Ask for step-by-step changes for safer refactors.

Good prompt template:

You are helping with a <language> project.
Task: <what to build/change>
Constraints: <rules, versions, style>
Output: <code/tests/explanation>

6) Practical Tutorial Exercise

Try this mini-project:

1. Create calculator.py.
2. Prompt Copilot:
   • “Create a Calculator class with add, subtract, multiply, divide methods and type hints.”
3. Prompt Copilot Chat:
   • “Generate pytest tests for edge cases, including divide by zero.”
4. Prompt Copilot Chat:
   • “Refactor for readability without changing behavior.”

You will practice generation, testing, and refactoring in one flow.

7) Best Practices

• Treat Copilot output as a draft, not final truth.
• Run tests and linters after accepting suggestions.
• Check security-sensitive code carefully (auth, SQL, shell commands).
• Prefer smaller prompts for precise edits.

8) Quarto Website Project (workflowr-like for Python)

This workspace now includes a Quarto website scaffold that mimics workflowr-style practice:

• source analysis files in analysis/
• reproducible execution with quarto render
• publishable static output in docs/

Project Files

• _quarto.yml: Quarto project and website configuration
• index.qmd: Website home page
• analysis/index.qmd: Analysis listing page
• analysis/01-data-check.qmd: Starter Python analysis report

Local Setup

1. Install Quarto: https://quarto.org/docs/get-started/
2. Create and activate a Python environment.
3. Install dependencies:

pip install -r requirements.txt

4. Render site:

quarto render

Rendered pages are written to docs/.

Publish

1. Run:

quarto render

2. Commit and push the rendered docs/ directory.
3. In GitHub repository settings, open Pages and set:
   • Source: Deploy from a branch
   • Branch: main
   • Folder: /docs

9) Troubleshooting

• No suggestions:
  • Check GitHub sign-in status in VS Code.
  • Ensure Copilot extensions are enabled.
  • Reload VS Code window.
• Poor suggestions:
  • Add more context and constraints.
  • Break tasks into smaller prompts.
• Slow responses:
  • Check internet connection and retry.

10) Next Steps

• Add project-specific prompt recipes to your team docs.
• Use Copilot to scaffold tests first, then implementation.
• Pair Copilot with CI checks for safe, fast iteration.

---

If you want, I can also generate a short version of this README for workshop handouts, or a team version with coding standards and review checklist.