Skip to content

JPeetz/agent-skills / skills/business-strategy/product-manager-toolkit

product-manager-toolkit

Verified

Use this skill when performing product management tasks like prioritizing features, synthesizing customer interviews, writing PRDs, setting OKRs, or managing stakeholder communication. Activates on "product management", "PM toolkit", "prioritization", "RICE", "PRD", "feature prioritization", "customer interview", "OKR planning", "stakeholder update", or "roadmap planning". Covers RICE prioritization, customer interview synthesis, PRD authoring, OKR setting with leading/lagging indicators, stakeholder communication templates, and decision frameworks.

FileKindSize
CHANGELOG.mdmarkdown2.0 KB
LICENSEother1.1 KB
SKILL.mdskill md17.2 KB
evals/evals.jsonconfig8.0 KB
references/sources.mdmarkdown5.4 KB
scripts/rice_scorer.pyscript8.2 KB

references/sources.md

View on GitHub
# Sources — Product Manager Toolkit

Provenance for the frameworks in `SKILL.md`, plus the reading they are condensed from.
Thresholds marked **(heuristic)** are defaults from practice, not published constants — argue with them.

---

## RICE

- **Origin:** Sean McBride, Intercom, 2016 — "RICE: Simple prioritization for product managers". Publicly documented; the four factors and the canonical impact multipliers (3 / 2 / 1 / 0.5 / 0.25) come from that post.
- **Confidence scale** (100 / 80 / 50 / 20) is from the same post; this skill adds the rule that unsupported items are capped at 50% rather than excluded.
- **Extensions:** `Reach × Impact × Confidence / Effort` variants (RICE+, weighted RICE, VAN) exist; all of them exist to separate evidence quality from value size. The skill keeps the base formula because it is the one a team can argue about in one meeting.
- **Known criticisms:** reach and effort are not commensurable, the score hides dependencies, and the confident team wins. The SKILL.md "Rules that keep RICE honest" subsection is derived from those critiques.

## Jobs to be Done / Interview synthesis

- Clayton Christensen, Taddy Hall, Karen Dillon, David Duncan — *Competing Against Luck* (2016). Job as progress-in-a-situation; "workaround = real competitor" follows directly.
- Bob Moesta & Chris Spiek — the Switch interview / forces diagram (push, pull, anxiety, habit). Underpins "ask about the last time, not hypotheticals".
- Rob Fitzpatrick — *The Mom Test* (2013). Three rules: talk about their life not your idea, ask about specifics in the past, talk less. The "never pitch" and "ask past behavior" guidance is this book.
- Teresa Torres — *Continuous Discovery Habits* (2021). Interview snapshots and opportunity solution trees; the participant-count discipline is adapted from her opportunity mapping.
- Affinity mapping: standard UX-research practice (KJ method, Jiro Kawakita, 1960s). The "cluster by behavior, not noun" rule is a practitioner convention, not from the original method.

## Opportunity Scoring

- Anthony Ulwick — *What Customers Want* (2005) and the Outcome-Driven Innovation method (Strategyn). Formula: `Importance + max(Importance − Satisfaction, 0)`.
- The scoring is only valid on *outcome statements* ("minimize the time to…"), not on features. Scoring features with it produces a wishlist.

## PRD practice

- Marty Cagan — *Inspired* (2nd ed., 2017): the product-discovery framing, problem before solution, and why a spec-style PRD fails.
- Atlassian / Aha! / Productboard public PRD templates contributed the artifacts-not-features structure and the must/should/won't split. The MoSCoW mnemonic is Dai Clegg, Oracle, 1994.
- Kill criteria and "revisit if" are adapted from Amazon's working-backwards habit and from Bezos's Type 1 / Type 2 decision framing (1997 shareholder letter, formalized 2015).

## OKRs

- John Doerr — *Measure What Matters* (2018); Andy Grove at Intel (1975) is the origin of the practice.
- Christina Wodtke — *Radical Focus* (2nd ed., 2021). The "grade the bet, not the execution" retro and the 0.7-is-good scoring norm come from here.
- Google's re:Work OKR guide for the 0–1 scoring convention and the "OKRs are not performance reviews" caveat.
- Leading vs lagging indicators: the distinction is standard in the 4 Disciplines of Execution (McChesney, Covey, Huling, 2012). The requirement that each objective carry at least one lagging KR is this skill's own guard against teams optimizing only the inputs.

## Decision frameworks

- **DACI:** developed at Intuit; widely taught at Atlassian. The "exactly one approver" rule is the whole point — consensus is a different, slower decision mode.
- **DARE:** Decision, Alternatives, Reasoning, Evidence — a decision-record convention (a lightweight cousin of ADRs, Michael Nygard, 2011). The added `Revisit if` line is this skill's contribution.
- **Type 1 / Type 2 decisions:** Jeff Bezos, 2015 Amazon shareholder letter — reversible decisions should be made fast by a small group; this is why the skill tells you to *skip* the framework when the call is cheap and reversible.

## Roadmap and stakeholder communication

- Cagan's four risks (value, usability, feasibility, business viability) — used implicitly in the PRD "Risks" table.
- Teresa Torres and Petra Wille on stakeholder alignment; the "bad news first, with a plan and a date" rule is standard crisis-communication practice applied to product updates.
- Patrick Lencioni — *The Five Dysfunctions of a Team* (2002), for the observation that trust dies on surprises rather than on misses.

---

## Reading list (order matters)

1. *The Mom Test* — Fitzpatrick. Two hours. Fixes interviews immediately.
2. *Inspired* — Cagan. The frame for everything else.
3. *Continuous Discovery Habits* — Torres. Turns discovery into a weekly habit.
4. *Measure What Matters* — Doerr. Skim; take the cadence, not the religion.
5. *Radical Focus* — Wodtke. The honest version of OKRs.
6. *Escaping the Build Trap* — Melissa Perri (2018). On outcomes vs outputs at org level.

## Source quality note

RICE and opportunity scoring are published formulas — cite them as such. OKR guidance is practice-derived and varies by organization; treat every threshold in this skill (0.7 is a good score, 8–12 interviews saturate, ≤50% confidence means research) as a heuristic to be tuned against your own data, not a standard to be enforced.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall