GitHub Copilot Lab Series

Learn to steer GitHub Copilot with intentional context, planning, tests, and governance guard rails.

Purpose

This repository is a set of hands-on labs designed to teach students (and teams) how to use GitHub Copilot responsibly and effectively within real-world code quality constraints. Each lab layers additional forms of context—global instructions, local artifacts, plans, tests, diagrams, decision logs—to demonstrate how they influence Copilot suggestions.

What You'll Learn

• Writing clear instruction files to shape style and architecture
• Planning-first prompting ("Do not write code yet") to reduce rework
• Test-first workflows that anchor Copilot's generations
• Refactoring duplication into small pure functions
• Reconciling conflicting constraints across docs, diagrams, and code
• Prompt snippet governance and lifecycle management
• Reflection-driven improvement habits

Guard Rails & Principles

We emphasize:

• Functional Core: Pure functions, no hidden mutable state
• Plan Before Code: Always request and refine a plan first
• Test-First & Diff Review: Tests define success; review generated diffs
• Minimal, Intentional Prompts: Short, directive, explicit output requests
• Governance & Maintenance: Snippet lifecycle (Active / Needs Review / Deprecated)
• Reflection: Every lab ends with a REFLECTION.md answered as questions

See CONTRIBUTING.md and .github/copilot-instructions.md for full guidelines.

Lab Overview

A quick summary (see labs-index.md for details):

Lab	Focus	Key Concepts
01	Basic Instructions	Global context shaping suggestions
02	Advanced Patterns	Structured prompts, refactoring targets
03	Integration Testing	Coverage gaps, quality gates
04	Refactoring Duplication	Pure helper extraction, DRY
05	Test-First	Tests constrain generation
06	Conflicting Constraints	Reconciliation strategies
07	Planning-First	Multi-step plan refinement
08	Snippet Governance	Lifecycle & metadata

Recommended Path

1. Start at Lab 01 and progress sequentially.
2. For each lab: Read README → Run a planning prompt → Capture the plan in PLAN.md → Implement → Reflect.
3. Apply global functional instructions consistently.
4. Use prompt snippets (see prompt-snippets/) for repeatable patterns.

Workflow Pattern

plan → clarify → implement → test → diff review → reflect → improve

Repository Structure

/.github/copilot-instructions.md   Global functional & style guidance
CONTRIBUTING.md                    Responsible Copilot usage guard rails
labs/                              Lab directories (each with README, PLAN, REFLECTION, src/)
labs-index.md                      Quick navigation index
prompt-snippets/                   Shared prompt snippet JSON + governance policy
slides/                            Intro & theory materials

Using Copilot Effectively

Step	Prompt Pattern	Goal
Plan	"Create a 5-step plan... Do not write code yet."	Elicit structure & ambiguities
Refine	"Clarify step 3; add edge cases."	Improve specificity
Implement	"Generate code per approved plan; keep pure functions."	Constrained generation
Test	"List edge cases for X; produce test skeletons."	Expand coverage
Review	"Explain diff impact in 3 bullets."	Understand changes
Reflect	"Summarize what worked and one improvement."	Continuous improvement

Decision & Conflict Handling

When artifacts disagree (instructions vs diagram vs code):

1. Identify conflicts explicitly
2. Propose reconciliation strategies (adapt docs, refactor code, hybrid compromise)
3. Implement a small, safe slice
4. Record rationale in DECISIONS.md

Snippet Governance

Prompt snippets include metadata comments (STATUS, LAST-REVIEW, OWNER). Keep them:

• Short
• Single-purpose
• Versioned via diffs
• Reviewed periodically

See labs/lab-08-prompt-snippet-governance/ and SNIPPET-GOVERNANCE.md for details.

Getting Started

Clone the repo, open a lab directory, and begin with the planning prompt.

git clone https://github.com/wouterfennis/copilot-lab.git
cd copilot-lab/labs/lab-01-basic-instructions
# Open README and start with a planning prompt in Copilot Chat

Contributing

Follow CONTRIBUTING.md. Always:

• Add or update a plan before large changes
• Keep functions small & pure when possible
• Update reflection questions after completing a lab

License

Licensed under the MIT License.

You are free to use, copy, modify, distribute, and sell copies of the software provided the copyright and permission notice are included.

Copyright (c) 2025 Wouter Fennis

---

Questions or suggestions? Open an issue or start a discussion.