Skip to content
Join free
deanpeters

Decision Rights Matrix skill

Define who decides, who is consulted, and who is told for key product decisions, with one accountable person each. Use when decisions stall, get reopened, or go to whoever pushes hardest.

$npx alyn-cli add decision-rights-matrix
View this skill on GitHubdeanpeters/Product-Manager-Skills/skills/decision-rights-matrix/SKILL.md
Pack stars
8,019
Pack installs
166k
Alyn Score
72/99
Updated
2 days ago

deanpeters

Decision Rights Matrix

What is the Decision Rights Matrix skill?

Decision Rights Matrix is a product management skill by deanpeters for Claude Code, Cursor, Codex and other AI coding agents. Define who decides, who is consulted, and who is told for key product decisions, with one accountable person each. Use when decisions stall, get reopened, or go to whoever pushes hardest.

Write down who decides what, for the product decisions that cause the most friction, so decisions stick. The output is a matrix of decisions against roles, exactly one accountable person per decision, the forums where decisions are made and how often, an escalation path, and a decision log.

When to use it

Use when decisions stall, get reopened, or go to whoever pushes hardest.

What the SKILL.md covers

  • Purpose
  • Input
  • Key Concepts
  • RACI and DACI: One for the Work, One for the Decision
  • One Accountable Person
  • Pick the Decisions That Hurt
  • Decide Vs. Input Vs. Veto
  • Forums: Where Decisions Get Made
  • Escalation
  • The Decision Log
Author
deanpeters
Pack
deanpeters/Product-Manager-Skills (30 skills)
Category
Product management, #93 of 185
Alyn Score
72 / 99
Files
4, includes references
License
Not stated
Last updated
2 days ago
Works with
Claude Code, Claude, Cursor, Codex, Gemini CLI, Copilot

Decision Rights Matrix compared to similar skills

SkillAlyn ScoreGitHub starsUpdated
Decision Rights Matrix · deanpeters728k2 days ago
Cs Product Manager · alirezarezvani6330k4 days ago
Product Decision Agent · mxyhi575168 days ago
Ansoff Matrix · phuryn7828k7 days ago

Install Decision Rights Matrix

Pick your agent. Every command is ready to paste.

$npx alyn-cli add decision-rights-matrix

Detects your agents (Claude Code, Cursor, Codex and 70+ more) and installs to the right folder. Add -g to install globally.

$npx skills add deanpeters/Product-Manager-Skills --skill decision-rights-matrix

Or use the open skills CLI directly.

01

Install it

Run the command above, or upload the folder in Claude.

02

Ask naturally

Describe the task. Your agent loads decision-rights-matrix when the request matches its description.

03

Name it if needed

If it does not trigger, say “use the decision-rights-matrix skill”. In Codex, type $decision-rights-matrix.

The skill itself

SKILL.md

FolderGitHub

Decision Rights Matrix

Purpose

Write down who decides what, for the product decisions that cause the most friction, so decisions stick. The output is a matrix of decisions against roles, exactly one accountable person per decision, the forums where decisions are made and how often, an escalation path, and a decision log.

The sentence to keep in your head: a decision with two deciders has none. When nobody knows who decides, decisions go to the loudest voice, get reopened by whoever lost, and leave the PM persuading instead of leading.

This is influence infrastructure for a matrix organization. It does not give the PM authority they don't have; it makes the authority that exists visible and agreed, which is what lets a PM lead without goodwill as the only tool.

Input

Works best with: The decisions that stall or get reopened, and who is involved in them.

Also useful: The org chart or teams involved, any existing forums (product review, steering committee), recent examples of a decision that went wrong, and who has senior sponsorship.

Anything supplied with the invocation itself (text after the skill name, a pasted context dump, or an appended ARGUMENTS: line) counts as answers already given. Use it and skip whatever it covers; don't re-ask.

Arriving empty-handed? That works too. The skill asks which decisions hurt most and who is involved, drafts a matrix as a proposal, and marks every role assignment as something the people named must agree, not assume.

Example invocations:

  • Sales overrides our list prices with special deals. Draft a matrix for pricing, roadmap, and launch go/no-go.
  • New platform team just joined. Who decides on architecture versus product scope?

Key Concepts

RACI and DACI: One for the Work, One for the Decision

  • RACI assigns the work: Responsible (does it), Accountable (owns the outcome, one person), Consulted (two-way input before), Informed (told after).
  • DACI frames a decision: Driver (runs the process, usually the PM), Approver (one person who decides), Contributors (supply input and evidence), Informed.

Use RACI to map recurring responsibilities across functions, and DACI for a specific decision you are about to make. Alternatives exist (RAPID, a "consent" model); the rule that survives all of them is one decider.

One Accountable Person

If two people are both accountable, the matrix hasn't decided anything. When the honest answer is "both," split the decision into two decisions (for example: what we build is the PM, how it's built is engineering). Keep splitting until each has one name.

Pick the Decisions That Hurt

Don't map everything. Start with the decisions where friction already exists, usually 5-8:

Decision Common conflict
What goes on the roadmap Sales requests vs strategy vs engineering capacity
Pricing and discount rules Sales vs product vs finance
Requirements and scope Product vs engineering
Launch go / no-go Product vs sales vs support vs legal
Portfolio entry or exit (including end of life) Product vs finance vs leadership
Customer-specific commitments Sales vs engineering

Add decisions that have caused a real fight in the last year. Leave out ones nobody argues about.

Decide Vs. Input Vs. Veto

Clarify three things people confuse: who decides, who must be consulted before (and what "consulted" means: their input is heard, not that they agree), and who can veto (rare, and only for a stated reason such as legal or safety). Most conflicts come from consulted people acting like approvers, or approvers deciding without consulting.

Forums: Where Decisions Get Made

A decision needs a place and a rhythm: a product review for roadmap and scope, a pricing council for discounts, a launch forum for go/no-go. For each: who attends, how often, what it decides (and what it doesn't), and what it needs ahead (a one-page brief). Too many forums become theater. Prefer fewer forums with clear mandates.

Escalation

When people disagree and the decider is unmoved, there must be a named next step: who the dispute goes to, within how many days, with what material. Without escalation, disputes simmer. Escalation is not failure; an unescalated conflict is.

The Decision Log

A short record of each significant decision: what, who decided, when, on what evidence, and what it supersedes. It ends relitigation ("we decided this in March, here's why") and teaches the next PM the reasoning.

Right-Size the Matrix

A small team needs a half-page: five decisions, one name each. A matrix organization with several business units needs forums, escalation, and a log. Over-building a matrix for a ten-person team is the same failure as under-building it for a global one. Offer a level and let the user move it.

Facilitation Source of Truth

Use workshop-facilitation as the interaction protocol when running this conversationally. Draft the matrix as a proposal and have the named people confirm it.

Default entry mode: Guided · set in pm-skills.config.yaml (library-wide default or a per-skill override); the user's own choice in the session always wins. On the first turn, offer all three modes, then open in the default without waiting for a pick: Guided asks the first question, Context dump invites the paste, Best guess asks for a one-sentence minimum and drafts with every assumption labeled.

Anti-Patterns (what this is NOT)

  • Not an org chart. It's about decisions, not reporting lines.
  • Not authority granted by the PM. Leadership has to agree it, or it's a wish.
  • Not a process for every decision. Only the ones that hurt.
  • Not stakeholder management. See stakeholder-mapping for the people dynamics.

Application

Use template.md for the scaffold.

Output convention. Artifact first: the decisions, the matrix (one accountable each), the forums, the escalation path, and the decision log format. Commentary short, after. Every role assignment is a proposal for the named person to confirm. Never present an assignment as agreed. Mark any name, cadence, or threshold you were not given as [NEEDS SOURCE].

If the PM brings a matrix, critique first. Does each decision have exactly one accountable person? Is "consulted" defined? Are forums and escalation named? Is there anything mapped that nobody argues about? Did the named people agree? Name gaps; offer rewrites last.

Step 1: Choose the Decisions

Ask which decisions stalled or got reopened. Pick 5-8. For each, name the recent example.

Step 2: Propose Roles

For each decision, propose the Driver or Responsible, the single Accountable or Approver, the Consulted, and the Informed. Where two people want to be accountable, split the decision.

Step 3: Define Forums and Escalation

For each forum: who, how often, what it decides, what it needs. Name the escalation path and its time limit.

Step 4: Set the Log

One line per significant decision, kept where the team can find it.

Step 5: Confirm With the People Named, and Hand Back

The matrix is a proposal until each accountable person agrees. Short summary (supplied, inferred, unknown), then a numbered ladder, evidence-favored first, labeled (Recommended): take it to the accountable people individually before a group review; get leadership to endorse it; start with the one decision that hurts most; or run a short trial and review in a quarter.


Examples

See examples/sample.md (Fieldlight discount overrides, SaaS) and examples/sample-industrial.md (Northfield launch go/no-go and pricing in a matrix organization).


Common Pitfalls

Pitfall 1: Two Accountable People

Symptom: "Product and Sales own pricing." Consequence: Nobody does, and the loudest wins. Fix: Split the decision until each has one name.

Pitfall 2: Consulted Acting as Approver

Symptom: A consulted team blocks a decision it can't make. Consequence: Decisions stall and resentment grows. Fix: Define consulted as input heard, not agreement.

Pitfall 3: Assignments Nobody Agreed

Symptom: The PM publishes a matrix. Consequence: It's ignored the first time it's inconvenient. Fix: The named people confirm it; leadership endorses it.

Pitfall 4: Mapping Everything

Symptom: A 40-row matrix. Consequence: Nobody reads it, and it hides the five that matter. Fix: Start with the decisions that already cause friction.

Pitfall 5: Forums Without Mandates

Symptom: A weekly meeting that reviews everything and decides nothing. Consequence: Theater, and a slower org. Fix: Each forum states what it decides and what it doesn't.

Pitfall 6: No Escalation Path

Symptom: A disagreement with no named next step. Consequence: It simmers until someone goes around the process. Fix: Name who it goes to, in how many days, with what material.

Pitfall 7: No Decision Log

Symptom: The same decision is reopened every quarter. Consequence: Time lost and trust eroded. Fix: A short log: what, who, when, why.


References

Skill Files

External Frameworks

  • RACI: standard responsibility assignment practice with no single originator.
  • DACI: the Driver, Approver, Contributors, Informed decision model, popularized at Intuit and widely used in product teams.
  • RAPID: Paul Rogers and Marcia Blenko, "Who Has the D?" Harvard Business Review (2006).

Provenance

Distilled from practitioner experience clarifying decision rights in matrix product organizations.

Tips for Decision Rights Matrix

  • • Start a fresh session after installing so your agent picks up the new skill.
  • • It bundles reference files. The agent reads them only when needed, so they cost no context until then.
  • • It is part of a pack with 30 skills. Install only the ones you need with --skill to keep triggers precise.
  • • If two skills overlap, mention the one you want by name.

Questions about Decision Rights Matrix

What does the Decision Rights Matrix skill do?

Define who decides, who is consulted, and who is told for key product decisions, with one accountable person each. Use when decisions stall, get reopened, or go to whoever pushes hardest.

How do I install Decision Rights Matrix?

Run "npx alyn-cli add decision-rights-matrix" in your project. The Alyn CLI installs it for Claude Code, Cursor, Codex and 70+ other agents. In Claude on the web, upload the skill folder as a ZIP under Customize > Skills.

Does Decision Rights Matrix work in Cursor and Codex?

Yes. Decision Rights Matrix uses the open SKILL.md format, so it works in Claude Code, Claude, Cursor, OpenAI Codex, Gemini CLI, GitHub Copilot and other agents that support agent skills.

When should I use the Decision Rights Matrix skill?

Use when decisions stall, get reopened, or go to whoever pushes hardest.

What is inside the Decision Rights Matrix skill?

The SKILL.md covers Purpose, Input, Key Concepts, RACI and DACI: One for the Work, One for the Decision, One Accountable Person, Pick the Decisions That Hurt, and the skill folder has 4 files.

What is a good alternative to Decision Rights Matrix?

Cs Product Manager by alirezarezvani is the highest-ranked similar skill from another creator, with an Alyn Score of 63 compared to 72 for Decision Rights Matrix.

Is Decision Rights Matrix free?

Decision Rights Matrix is open source on GitHub. It is free to install; check the license before commercial use.

Who made Decision Rights Matrix?

Decision Rights Matrix is published by deanpeters in the deanpeters/Product-Manager-Skills repository, together with 29 other skills.

Similar skills. From other creators.