P3. GitHub Copilot for scripts

Outcome: you can use GitHub Copilot to write, explain and test lab scripts, and you review every suggestion before you keep it.

Before you start: S0 (command line and Git), a GitHub account and VS Code. Copilot Free is enough for the labs.

Skills: inline suggestions, Copilot Chat, inline chat and slash commands in VS Code; agent mode; the Copilot CLI; prompt and context crafting; repository custom instructions; generating tests and documentation; privacy settings, matching public code and content exclusions (on business plans); responsible use.

Courses: GH-300T00: GitHub Copilot Fundamentals parts 1 and 2.

Labs

1. Set up: install VS Code, sign in with GitHub, turn on Copilot Free and open your lab scripts repo (from S0, or from C2 if you did it).
2. PowerShell: write a function with inline suggestions, then use chat to add comment-based help and Pester tests. Run them.
3. Bash: ask Copilot to explain an unfamiliar script line by line, and check its answer against the man pages.
4. Context: add repository custom instructions with your lab rules (VM names, "lab only, never production", how to test). Compare the suggestions before and after.
5. Privacy: read your Copilot settings, including suggestions that match public code, and write down what content exclusions would add on a business plan.
6. Review log: for every suggestion you kept, note what you checked and what you changed. If you did C2, do one task with both tools and compare.

Prove it: the repo with passing tests and the custom instructions file, plus the review log. A partner runs your scripts from the README alone.

Certs: GH-300 (use GitHub Copilot responsibly 15–20%, GitHub Copilot features 25–30%, data and architecture 10–15%, prompt engineering and context crafting 10–15%, developer productivity 10–15%, privacy, content exclusions and safeguards 10–15%).

Gate: GH-300 happens here, with one review week. Take the free practice assessment on Microsoft Learn first.