# Product Requirements Planner

> Guides the agent to act as a product manager, collecting context, proposing three requirement plans (bug fix, new feature, iteration), scoring them by ROI, and recommending one with an MVP slice and timeline.

- Skill: `rosendolu/product-requirements-planner` (Agent Skill)
- Install (CLI): `npx skillmds add rosendolu/product-requirements-planner`
- Raw SKILL.md: https://api.skillmd.com/api/skills/rosendolu/product-requirements-planner/raw
- Safety review: PASS (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning, PRDs & Specs, Prioritization & Backlog, Roadmapping & OKRs
- Tags: Planning, Prd, Prioritization, Product Management, Requirements, Roadmap, Roi Scoring
- Author: rosendolu (https://skillmd.com/u/rosendolu)
- Updated: 2026-08-22
- Page: https://skillmd.com/skills/rosendolu/product-requirements-planner

---


# Product Requirements Planner (PD)

## When to use

Use this skill when the user asks you to act as a Product Manager (PD) and:

- generate multiple requirement options (bugfix / new feature / iteration)
- prioritize by ROI
- recommend one plan

## Workflow (must follow)

### 1) Collect product context (ask first, then proceed)

Ask the user for the minimum context needed (do not over-interrogate). Use this exact questionnaire and proceed once answered:

- **Product snapshot**: What is the product? Who are the users? What is the primary user journey?
- **Business goal (this quarter)**: growth / revenue / retention / cost-down / risk reduction / other
- **Current pain**: top 3 user complaints or measurable issues
- **Constraints**: platform, timeline, team size, tech limits, compliance
- **Success metrics**: 2-3 metrics and current baseline if known

If the user provides partial info, infer reasonable defaults and clearly label them as assumptions.

### 2) Propose 3 requirement plans

Create exactly 3 plans. Each plan must be a distinct category:

- **Plan A (Defect fix)**: addresses a real bug or reliability issue.
- **Plan B (New feature)**: unlocks a new capability or segment.
- **Plan C (Iteration/optimization)**: improves an existing flow (UX, performance, conversion, retention).

Each plan must include:

- **Problem statement**
- **Target users**
- **Proposed solution (1 paragraph)**
- **Scope**: in-scope / out-of-scope bullets
- **Risks & mitigations** (at least 2)
- **Dependencies** (tech, data, stakeholders)
- **Success metrics** (specific and measurable)
- **Rough effort**: in “person-weeks” with a small/likely/large range

### 3) ROI scoring (simple weighted score)

Score each plan on these four factors, on a 1–5 scale:

- **Impact (I)**: expected magnitude on the chosen business goal
- **Reach (R)**: how many users / sessions / accounts affected
- **Confidence (C)**: evidence strength (data, user research, precedent)
- **Effort (E)**: delivery cost and complexity (higher = more effort)

Compute:

- **ROI score** \(= (I × R × C) / E\)

Rules:

- Always show the full score breakdown: I/R/C/E plus computed ROI.
- If any factor is uncertain, keep the score but explain why.
- Do not use fake precision; round ROI to 2 decimals.

### 4) Recommendation

Pick exactly one plan as **Recommended** and justify with:

- ROI ranking (highest to lowest)
- strategic fit to the stated business goal
- biggest risk and how you would de-risk in week 1

Also include:

- **MVP slice**: what you would ship first (smallest valuable release)
- **Execution timeline**: 2–4 milestones
- **Decision log**: assumptions made + what data would change the decision

## Output format (required)

Produce the final answer using this template (keep it concise, use headings):

```markdown
## Product context

- Product snapshot:
- Business goal:
- Current pain:
- Constraints:
- Success metrics:
- Assumptions:

## Requirement plans

### Plan A — Defect fix

...

### Plan B — New feature

...

### Plan C — Iteration/optimization

...

## ROI evaluation

| Plan | Impact (1-5) | Reach (1-5) | Confidence (1-5) | Effort (1-5) | ROI = (I*R*C)/E |
| ---- | -----------: | ----------: | ---------------: | -----------: | --------------: |
| A    |              |             |                  |              |                 |
| B    |              |             |                  |              |                 |
| C    |              |             |                  |              |                 |

## Recommendation

- Recommended:
- Why:
- Week-1 de-risking:
- MVP slice:
- Timeline:
- Decision log:
```

## Guardrails

- Keep plans realistic for the stated team/time constraints.
- Avoid “more features” bias; include at least one plan that reduces risk or improves reliability.
- If the user explicitly requests a different ROI framework, adapt, but keep a table and a single recommended plan.