GH-300: GitHub Copilot

Note: This lab pre-provisions a Windows VM and GitHub Enterprise sign-in credentials at start — allow up to 10 minutes for the environment to become ready before beginning the exercises.

Follow Alex Kim through a complete Copilot Foundations lab on a NorthPeak Robotics workstation. In Exercise 1, sign in to skillmeupai.ghe.com, install and authenticate Copilot in both VS Code and Visual Studio 2026, practice the four ghost-text keyboard shortcuts on fleet-utils.js, drive Copilot Chat with slash commands and @workspace, and lock down fleet-secrets/ with a .copilotignore. In Exercise 2, evaluate Copilot suggestions on fleet-auth.js for five planted security vulnerabilities and score Copilot's fixes against the three-layer validation stack. In Exercise 3, set the user-level duplicate-detection toggle to Block, apply the attribution decision tree, and commit NorthPeak's responsible-use notes to .github/COPILOT_USAGE.md. By the end you will have the full Day-1 Copilot operating pattern plus the responsible-AI discipline the course teaches.

Note: This lab pre-provisions a Windows VM and GitHub Enterprise sign-in credentials at start — allow up to 10 minutes for the environment to become ready before beginning the exercises.

Hands-on practice across all six prompting + context levers from the lesson: compare zero-shot vs few-shot prompting side by side; author a repo-wide .github/copilot-instructions.md, a scoped .github/instructions/typescript.instructions.md, and a shareable .github/prompts/review-pr.prompt.md the squad invokes with /review-pr; stand up a FleetBoard Core Copilot Space on skillmeupai.ghe.com with four pinned sources and feel the grounding delta against plain Chat; and prototype a FleetBoard operator dashboard from an empty folder with Copilot Agent mode for Michael's Thursday leadership demo. By the end of this lab you will be able to pick the right Copilot tool for each shape of task and codify repeat winners as team-wide assets.

Note: This lab pre-provisions a Windows VM and GitHub Enterprise sign-in credentials at start — allow up to 10 minutes for the environment to become ready before beginning the exercises.

Three hands-on exercises on Copilot's two surfaces outside the editor. In Exercise 1, sign in to skillmeupai.ghe.com, install GitHub Copilot CLI with winget, sign it in with copilot login --host, and ask it to suggest, explain, and revise PowerShell, gh, and git commands, use copilot -p for one-shot explanations, and decline a mutating command in the tool approval prompt. In Exercise 2, open multi-turn interactive Copilot CLI sessions, iterate on a FleetBoard log-triage pipeline across several prompts, then promote the final one-liner to a parameterized, committed fleet-ops/log-triage.ps1. In Exercise 3, push a FleetBoard feature branch to GitHub Enterprise, generate a Copilot PR summary from the diff, request a Copilot Code Review, tune the review with a repo-level .github/instructions/review-standards.instructions.md, then (where the tier allows) assign @copilot to a FleetBoard issue so Copilot cloud agent lands a draft PR. All work runs on the lab's Windows 11 VM; the GHE repos the student creates are wiped at lab end.

Note: This lab pre-provisions a Windows VM and GitHub Enterprise sign-in credentials at start — allow up to 10 minutes for the environment to become ready before beginning the exercises.

Four exercises across two FleetBoard starter projects on the same Windows 11 VM. Exercises 1–2 drill the Copilot chat modes (Ask / Edit / Agent) and the Agent Mode extensions (MCP server registration, Sub-Agent delegation) against C:\labs\copilot-agent-mode\. Exercises 3–4 drill the daily productivity loop (refactor, document, generate a realistic fixture, test with /tests, OWASP-grounded security review, input validation) against a parallel task-scheduler module in C:\labs\copilot-productivity\. By the end of this lab you will have hands-on counterparts to every topic in the paired lesson — three modes, MCP, Sub-Agents, refactoring, documentation, testing, and security review.

Note: This lab pre-provisions a Windows VM and GitHub Enterprise sign-in credentials at start — allow up to 10 minutes for the environment to become ready before beginning the exercises.

Configure the policy surfaces that keep Copilot inside the NorthPeak Robotics data perimeter: write and test content-exclusion patterns against FleetBoard's secrets paths, trigger and inspect duplication detection on a public-code match, review Copilot audit log events for a window of org activity, and tune organization-level policies to match the rollout plan. By the end of this lab you will be able to configure content exclusions, interpret audit log events, and set the org-level posture Copilot runs under.