cagan-brain

name

cagan-brain

description

Complete knowledge base for Marty Cagan's product canon — INSPIRED, EMPOWERED, and TRANSFORMED — all three books in one skill, automatically routed to the right book and chapter for any question. Use this skill whenever the user asks anything about product management, product discovery, the four risks, empowered product teams, feature teams, OKRs or team objectives, prototyping, customer interviews, product vision or strategy, coaching or one-on-ones, staffing or hiring product people, team topology, stakeholder management, the product operating model, or company transformation — even if they never mention Cagan, a book title, or these frameworks by name. If the user mentions Cagan, SVPG, Silicon Valley Product Group, INSPIRED, EMPOWERED, or TRANSFORMED, this is the skill to use.

Cagan Brain

One entry point to the complete Cagan canon: INSPIRED, EMPOWERED, and TRANSFORMED. Your job is to route every question to the right book(s), read the right chapter(s), and answer with citations — the user should never have to know which book covers what.

The three books at a glance

Skill / book	Scope	When it's the primary source
cagan-inspired (INSPIRED)	Team-level product work: discovery, prototyping, roles, OKR technique, product culture	"How do we build / discover / validate the product?"
cagan-empowered (EMPOWERED)	The leadership layer: coaching, staffing, vision, strategy, team objectives, topology	"How do I lead / coach / staff / organize the people?"
cagan-transformed (TRANSFORMED)	Organization-wide change to the product operating model	"How do we change the whole company to work this way?"

The books build on each other in a strict order: INSPIRED → EMPOWERED → TRANSFORMED. EMPOWERED assumes INSPIRED's discovery model (its coaching topics reference INSPIRED techniques), and TRANSFORMED assumes both — it summarizes discovery in ch17 and the leadership layer in ch13. Route from the most specific book outward, and surface the dependency when it matters.

Routing algorithm

1. Parse the intent. Decide which layer the question lives in: (a) how teams do product work, (b) how leaders develop and lead people, or (c) how the whole organization changes. Many questions live in two or three layers at once — that's normal and the answer should come from multiple books.

2. Consult routing.md (in this skill's directory). Find the matching topic rows. Every row lists the book(s) and their chapter numbers. When a topic spans books, the row already says so — don't pick one book and ignore the others.

3. Read each candidate book's SKILL.md before answering. The three companion skills ship in the same directory as this one: ../cagan-inspired/SKILL.md, ../cagan-empowered/SKILL.md, ../cagan-transformed/SKILL.md (relative to this skill's folder). Each SKILL.md holds the Core Frameworks summary, the full chapter index, and a topic index — use the topic index to confirm or expand the chapter list from step 2. If you can't resolve the sibling paths directly, search for cagan-inspired/SKILL.md with your file tools.

4. Read the chapter files. E.g. ../cagan-inspired/chapters/ch33-principles-discovery.md. The Core Frameworks in the SKILL.mds often cover 80% of a question; read full chapters whenever the user wants specifics — techniques, numbers, thresholds, anti-patterns. Follow the "Connects To" sections inside chapters: they point at chapters in the other books that continue the same idea, which is exactly how you find the second and third book for a cross-book answer.

5. Answer with citations (format below). Never ask the user which book they mean — routing is your job. Only stop short when the question is genuinely outside the canon (not about product/leadership/transformation as Cagan frames them); then say so plainly and don't improvise from general knowledge.

The point of reading the files instead of answering from memory is accuracy and attribution: every claim in your answer must trace to a real chapter, and the user should be able to open the book and verify.

Answer format

• Single book. Answer normally, citing inline at each substantive claim: (INSPIRED, ch33 — Principles of Product Discovery). Include the chapter title when you can — it makes the citation scannable.

• Multiple books. Organize the answer by theme, and label each part with its source:

  • "On empowerment — EMPOWERED, ch04 (Empowered Product Teams): ..."
  • "On the discovery work itself — INSPIRED, ch33 (Principles of Product Discovery): ..."
  • "On making the change org-wide — TRANSFORMED, ch30–32 (Transformation Tactics): ..." Then close with a short Across the books synthesis: how the pieces fit together, which book goes deepest on the topic, and what to read first for more.

• When the books differ in emphasis, present both views explicitly rather than averaging them. Example: vision is the team's context in INSPIRED (ch24–25), but EMPOWERED (ch37–38) adds the leader's creation-and-ownership role and the visiontype. Name both framings and the reader can pick the lens they need.

Common question shapes

• "What's the difference between X and Y?" — route each concept, then compare across books and chapters.
• "Which book should I read?" — describe each book's scope (table above) and where the question sits in the canon.
• Vague questions ("how do we do product right?") — answer with the fundamentals across all three books and point to the chapters for depth, rather than asking for clarification.
• Situation questions ("my PMs report to sales") — route to the relevant chapters and apply the frameworks to their situation, with citations.

Supporting files

• routing.md — the cross-book topic map (topic → book + chapters). Read it in step 2.
• Each companion skill also carries glossary.md, patterns.md, and cheatsheet.md; point the user to the relevant one for follow-up depth (e.g., "for the full decision rules see the EMPOWERED cheatsheet").