End-to-End Demo (20 min)

This guide walks through a complete Copilot showcase, hitting every major capability in a natural flow. Each section includes the prompt, what to narrate, and timing guidance.

Sequence Overview

1. Plan a New Service

Narrate: “Let’s start by asking Copilot to plan a new feature for our platform — a wishlist service.”

Create an implementation plan for adding a `wishlist-service` to this monorepo.
Keep it consistent with existing patterns in `services/*` (Express + TypeScript),
include API endpoints, database schema ideas, events, tests, Docker, Helm, and CI
updates. Output phased steps with clear acceptance criteria.

What to highlight: Copilot understands the existing architecture and generates a plan that follows established conventions.

Transition: “Now that we have a plan, let’s have Copilot start implementing it.”

2. Scaffold the Implementation

Narrate: “The coding agent can take our plan and produce working code that matches the project’s patterns.”

Implement health/readiness endpoints for `services/payment-service` and add
unit tests for them, following existing code style and API response format.

What to highlight: The generated code follows the project’s Express patterns, TypeScript types, and test conventions.

Transition: “Let’s see how Copilot reviews the code it just generated.”

3. Review the Changes

Narrate: “Copilot code review examines code for correctness, security, and consistency — just like a senior engineer would.”

Focus this review on API contract consistency (`success/data/error/meta` shape),
error handling, and TypeScript strictness across touched files.

What to highlight: Scoped reviews produce deeper, more actionable findings than broad reviews.

Transition: “Now let’s bring in our custom agents for specialized tasks.”

4. Generate BDD Tests

Narrate: “We’ve configured a BDD specialist agent that understands Gherkin and Playwright.”

@bdd-specialist Create a BDD coverage matrix and Gherkin feature file for the
checkout flow in `services/order-service`. Include happy path, edge cases
(empty cart, out-of-stock), error scenarios (payment failure), and
accessibility checks.

What to highlight: The agent generates structured test scenarios with proper Given/When/Then format, covering edge cases specific to the checkout flow.

5. Create CI Workflow

Narrate: “Our GitHub Actions expert agent designs secure, least-privilege CI pipelines.”

@github-actions-expert Create a reusable CI workflow for backend services
that runs lint, unit tests, and container image build — triggered only when
that service's files change. Follow least-privilege permissions and pin
all actions.

What to highlight: Actions pinned to SHA, minimal permissions, path-based triggers.

6. Visualize Architecture

Narrate: “Agent skills can produce artifacts like diagrams from natural language.”

Create an Excalidraw architecture diagram showing the Octo E-Shop
microservices: frontend, user-service, product-service, cart-service,
order-service, payment-service, and their communication paths.

What to highlight: The output is a real .excalidraw file that opens directly in the Excalidraw editor.

7. Terminal Assistance

Narrate: “Copilot CLI brings AI assistance right into the terminal.”

Propose and apply a minimal change that improves developer experience in this
repo. Show the diff and rationale.

What to highlight: Copilot CLI can not only suggest changes but actually apply them, creating a real diff.

8. Code Quality & Security

Narrate: “Finally, let’s look at GitHub’s built-in code quality features enhanced by Copilot.”

What to highlight: Code quality findings surface directly in the developer workflow — no context switching needed.

Recovery Tips