Capstone: Reviewed Multi-Agent Automation

In this capstone you will build a reviewed multi-agent automation with explicit guardrails, a human approval point, an observable run trace, and a written account of what it must not do unsupervised.

Units

1. Unit 10.00: Choosing a workflow that justifies a crew
2. Unit 10.01: Building it with guardrails from the start
3. Unit 10.02: Instrumenting the run for review
4. Unit 10.03: Testing the failure paths deliberately
5. Unit 10.04: The handover note and its stated limits

Module work

• Activity
• Knowledge check
• Summary