Agentic isn't binary. There are five useful rungs, and choosing your rung is the single most important product decision in this course. Each one is a different product with different risks, different costs and different users.

Rung	What the System Does	Who Acts	Risk
1. Suggests	Recommends what you could do	The human, entirely	Almost none
2. Drafts	Prepares the action, doesn't send it	The human clicks send	Low, wasted time
3. Acts with approval	Chooses and prepares; waits for a yes	The human approves each one	Moderate, depends on approval quality
4. Acts within limits	Does it, within pre-set boundaries	The system, bounded	Real, bounded by the limits you set
5. Acts freely	Does whatever it judges necessary	The system	High and hard to bound

Most Products Should Ship at Rung 2 or 3

There is enormous pressure, internal and competitive, to build rung 4 or 5. Resist it for the first release. The value gap between rung 2 and rung 4 is usually smaller than teams expect, and the risk gap is much larger. Ship at 2 or 3, earn the data, and move up deliberately once you know your failure modes.

Agentic AI and MCP for Product Managers

Module 1What 'Agentic' Actually Means

Module 2How an Agent Actually Works

Module 3MCP: What It Is and Why You Should Care

Module 4Designing the Tool Layer

Module 5Permissions and the Human in the Loop

Module 6Evaluating Agents

Module 7Cost, Latency and Unpredictability

Module 8Designing What the User Sees

Module 9Risk and Blast Radius

Module 10When an Agent Is the Wrong Answer

Module 11Final Project: Design One Real Agent End to End

Module 12Appendix: The Prompt Library

The Autonomy Ladder