Exercise 05: Document Use Cases

Objective

Portfolio exercise. Document 3 AI use cases demonstrating when and why to use multi-agent pipelines.

Required Reading

• Foundation README -- "Why multi-agent?" section
• Agents | Cursor Learn -- Overview of what agents are and how they work
• Working with Agents | Cursor Learn -- Practical patterns for agent-assisted development
• Putting It Together | Cursor Learn -- End-to-end workflows that demonstrate multi-agent coordination

CursorClaude Code

The exercises and validation below work in Cursor. Use the Cursor documentation links in Required Reading.

The use cases you document here are transferable. Whether using Cursor's subagent dispatch or Claude Code's sequential prompting, the reasoning about when and why to split work across agents (specialization, safety, cost control) is the same. The agent mapping you define would use the same role names in either system.

Scenarios

1. Add pagination to a REST API (2 files, straightforward)
2. Migrate a PostgreSQL database schema with zero downtime (5+ files, cross-cutting)
3. Add OAuth2 integration with Google and GitHub providers (6+ files, auth domain, security-critical)

Task

For each scenario, write: - Task description: What needs to be done (2-3 sentences) - Agent mapping: Which agents handle which parts - Artifacts produced: List the pipeline artifacts that would be created - Why multi-agent: 2-3 sentences explaining why splitting this across agents is better than a single agent

Output

Write to docs/foundation/tutorials/outputs/05-use-cases.md with headings ## Use Case 1, ## Use Case 2, ## Use Case 3, each with the 4 subsections above.

Validation

python3 docs/foundation/tutorials/verify.py --exercise 05

Checks: 3 use cases present, each has all 4 subsections, "Why multi-agent" sections are at least 20 words.

Answer

This is a portfolio exercise with no single correct answer. A strong response:

• Task description: 2-3 sentences specifying scope, file count, and domains touched
• Agent mapping: Uses correct agent names (jg-subplanner, jg-worker, etc.) and explains what each does for this specific task
• Artifacts produced: Lists all pipeline artifacts including conditional ones (e.g., debug-diagnosis.json if tests fail)
• Why multi-agent: Explains the specific benefit for THIS task -- separation of concerns, dedicated review for domain-specific issues, traceable handoffs

Common mistakes: generic "Why multi-agent" sections that could apply to any task, omitting the debugger from complex scenarios, not distinguishing standard from high-tier agents for security-critical work.

See docs/foundation/tutorials/solutions/05-use-cases-guide.md in the source repo for a complete exemplar.