Skip to content

Use Cases

Concrete scenarios showing how the Copilot Expert Hub is used in practice.

Build a New Microservice

Situation: You're starting a new backend project — an order management system with event-driven architecture.

Agent Chain:

graph LR
    A[Task Orchestrator] --> B[Lead Architect]
    B --> C[Python Expert]
    C --> D[Test Strategist]
    D --> E[Code Reviewer]
    E --> F[DevOps Agent]

Steps:

  1. Task Orchestrator breaks the project into epics and tasks
  2. Lead Architect designs the architecture with ADRs and C4 diagrams
  3. Python Expert implements domain models, services, and API endpoints
  4. Test Strategist defines the test strategy and implements tests
  5. Code Reviewer reviews code quality, security, and best practices
  6. DevOps Agent creates Dockerfile, CI/CD pipeline, and deployment

Example Prompt:

@workspace Use the Lead Architect agent to design an event-driven
architecture for an order management system.
Use DDD, Hexagonal Architecture, and Pub/Sub for events.

Refactor Legacy Code

Situation: A mature Python project has accumulated technical debt — no tests, coupled modules, unclear architecture.

Agent Chain:

graph LR
    A[Architecture Reviewer] --> B[Context Manager]
    B --> C[Python Expert]
    C --> D[Test Strategist]
    D --> E[Code Reviewer]

Steps:

  1. Architecture Reviewer analyzes the existing architecture and identifies risks
  2. Context Manager documents the current state and decisions
  3. Python Expert refactors step-by-step using Repository Pattern, DI, and Clean Architecture
  4. Test Strategist adds tests — characterization tests first, then unit tests
  5. Code Reviewer ensures each refactoring step improves quality

Example Prompt:

@workspace Use the Architecture Reviewer agent. Analyze the
existing codebase and create a refactoring plan prioritized
by risk and effort.

Prepare a Sprint Review

Situation: End of sprint — you need to present progress to the client.

Agent Chain:

graph LR
    A[Context Manager] --> B[Stakeholder Agent]
    B --> C[Presentation Agent]

Steps:

  1. Context Manager gathers all relevant changes, PRs, and decisions
  2. Stakeholder Agent translates technical details into business language with KPIs
  3. Presentation Agent creates a Marp slide deck using the Project Review template

Example Prompt:

@workspace Use the Stakeholder Agent. Create a sprint review
report for Sprint 4. Focus on business impact and demo-ready
features.

Plan a Cloud Migration

Situation: An on-premise system needs to be migrated to Google Cloud Platform.

Agent Chain:

graph LR
    A[Lead Architect] --> B[GCP Architect]
    B --> C[DevOps Agent]
    C --> D[Context Manager]

Steps:

  1. Lead Architect designs the target architecture (Cloud-Native, 12-Factor)
  2. GCP Architect selects the appropriate GCP services (Cloud Run, Cloud SQL, Pub/Sub)
  3. DevOps Agent creates Terraform modules, CI/CD pipelines, and monitoring
  4. Context Manager documents all migration decisions as ADRs

Example Prompt:

@workspace Use the GCP Architect agent. Design a cloud-native
architecture on GCP for our existing monolith backend.
Use Cloud Run, Cloud SQL, and Pub/Sub.

Onboard a New Team Member

Situation: A junior developer is joining the team and needs to become productive quickly.

Agent Chain:

graph LR
    A[Context Manager] --> B[Code Reviewer]
    B --> C[Python Expert]

Steps:

  1. Context Manager provides project context, architecture overview, and ADRs
  2. Code Reviewer gives constructive feedback on initial PRs
  3. Python Expert explains patterns, conventions, and best practices in the project context

Example Prompt:

@workspace Use the Context Manager agent. Give me a complete
overview of the project: architecture, tech stack, key decisions,
and conventions.

Create a Project Proposal

Situation: You need to create a technical proposal for a new client.

Agent Chain:

graph LR
    A[Lead Architect] --> B["Proposal/Pitch Agent"]
    B --> C[Presentation Agent]

Steps:

  1. Lead Architect outlines the technical solution and feasibility
  2. Proposal/Pitch Agent creates a roadmap, milestones, and effort estimates
  3. Presentation Agent packages everything into a Client Pitch Marp deck

Example Prompt:

@workspace Use the Proposal/Pitch Agent. Create a project
proposal for an IoT monitoring platform with a 3-month
roadmap and milestones.

View All Agents Get Started