Hands-on GitHub Copilot labs

[!TIP] Download only what you need: the learner-kit catalog includes a ZIP for each of the 21 exercise variants, with instructions, images, tests, synthetic data, licenses and integrity checks. Preparation guides reuse the associated exercise kit rather than duplicating it.

Use these numbered exercises for focused professional scenarios. They share the repository’s evidence-first approach but are not fantasy adventures. Durations are facilitation estimates, not measured completion guarantees.

[!IMPORTANT] Work on one copied fixture at a time. Keep temporary files and caches on the selected work drive; do not run every build or profiler on a shared workstation. Start with environment setup.

Preparation

Lab	Outcome
C# environment	SDK/target agreement, one build and actual test discovery
Python environment	Interpreter, import root and isolated testing
Copilot access	Account, target and harmless request evidence
Spec Kit setup	Pinned skills integration without overwriting existing work
SDK setup	Offline logic separated from authenticated inference

Core workflow

Choose a language path; do not build the C# and Python copies simultaneously.

Lab	C#	Python	Outcome
01	Context and interface	Shared exercise	Roles, harnesses, evidence and permissions
02	Analyze/document	Analyze/document	Source-grounded onboarding
03	Develop availability	Develop availability	End-to-end book-copy availability
04	xUnit tests	pytest tests	Tests that detect wrong behavior
05	Refactor	Refactor	Characterization and semantic preservation

Engineering and collaboration

Lab	Focus
06 - Shopping prototype	Small PRD, integer-cents domain and accessible UI checks
07 - Duplication	Shared mechanics versus distinct business rules
08 - Large functions	Error paths, inventory compensation and extraction
09 - Complex conditions	Decision tables and equality boundaries
10 - Profiling	Small measurements, controlled scope, no promised speedup
11 - Issues	Local reproduction to reviewed change
12 - Secret remediation	Revocation-first simulation without real credentials

Spec-driven development and agent customization

Lab	Case and stack
13 - Greenfield	RSS subscription contract; Node/TypeScript with .NET/Python/Go adaptations
14 - Brownfield feature	Existing dashboard; owner-scoped metadata and compatibility
15 - Customization	Instructions, prompts, skills and agent handoffs
16 - Copilot SDK	Read-only support tool, offline contracts and optional live inference
17 - Modernization	Python CSV to SQLite; reconciliation, non-overwrite and rollback

How to choose a Spec Kit case

---
config:
  theme: base
  look: classic
  themeVariables:
    darkMode: false
    background: "#ffffff"
    primaryColor: "#f5f5f5"
    primaryTextColor: "#111111"
    primaryBorderColor: "#555555"
    secondaryColor: "#e0e0e0"
    secondaryTextColor: "#111111"
    secondaryBorderColor: "#666666"
    tertiaryColor: "#bdbdbd"
    tertiaryTextColor: "#111111"
    tertiaryBorderColor: "#444444"
    lineColor: "#444444"
    textColor: "#111111"
    mainBkg: "#f5f5f5"
    nodeBorder: "#555555"
    clusterBkg: "#ffffff"
    clusterBorder: "#999999"
    edgeLabelBackground: "#ffffff"
    actorBkg: "#e0e0e0"
    actorBorder: "#555555"
    actorTextColor: "#111111"
    actorLineColor: "#777777"
    signalColor: "#333333"
    signalTextColor: "#111111"
    labelBoxBkgColor: "#f5f5f5"
    labelBoxBorderColor: "#777777"
    labelTextColor: "#111111"
    loopTextColor: "#111111"
    activationBkgColor: "#bdbdbd"
    activationBorderColor: "#555555"
    noteBkgColor: "#f5f5f5"
    noteTextColor: "#111111"
    noteBorderColor: "#777777"
    attributeBackgroundColorOdd: "#f5f5f5"
    attributeBackgroundColorEven: "#e0e0e0"
---
flowchart TD
    accTitle: Choose a spec-driven development case
    accDescr: Distinguish a new product, a new feature in an existing product, and a technical modernization that preserves consumer behavior.
    Start{"Does working behavior already exist?"}
    Start -->|No| Greenfield["13: Specify a small new product"]
    Start -->|Yes| Change{"What must change?"}
    Change -->|New user behavior| Feature["14: Preserve baseline and add a feature"]
    Change -->|Technical boundary| Modernize["17: Preserve contracts and migrate data"]
    Greenfield --> Evidence["Requirements, implementation and checks"]
    Feature --> Evidence
    Modernize --> Evidence

Legend. Diamonds are scope decisions, rectangles are exercise paths, and solid arrows show selection flow. Labels distinguish new behavior from a storage change.

Explanation. All three paths use Spec Kit artifacts, but their starting evidence and failure risks differ. Modernization requires preserved output, reconciled data, and rollback, not just a new implementation.