# Multica Review Product

> Requirements/PRD dedicated review framework. Called by ProductReviewer to professionally analyze ProductManager's PRD (scope clear/goal explicit/acceptance testable/constraints complete/open items), output PASS/FAIL + fix list, report to Leader.

- Skill: `it235/multica-review-product` (Agent Skill)
- Install (CLI): `npx skillmds@latest add it235/multica-review-product`
- Raw SKILL.md: https://api.skillmd.com/api/skills/it235/multica-review-product/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: it235 (https://skillmd.com/u/it235)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/it235/multica-review-product

---


# Requirements (PRD) Professional Review (ProductReviewer)

Structured professional review framework for **PRD / requirement artifacts**. Called by `ProductReviewer`; reviews the PRD link `ProductManager` returns via `multica-pm-artifact-publish`.

## When to use
- ProductReviewer receives a "review PRD" dispatch from Leader.
- Entering a re-review round after PRD changes (check previous round's fix list item by item).

## Review dimensions (conclusion per item)
1. **Scope clear**: boundaries unambiguous, splittable, no implied scope.
2. **Goal explicit**: problem measurable, no vagueness.
3. **Acceptance testable**: every AC- objectively verifiable, no "good experience" vagueness.
4. **Constraints complete**: permissions, exceptions, compliance, dependencies, data口径 listed.
5. **Open items**: OP- fully listed, not blocking downstream dev (blocking OP- = blocking item).

## Output format
```
【PRD Review】<PRD link>
Conclusion: PASS / FAIL
Blocking items (required on FAIL, each: rationale / involved point / fix direction):
- ...
Suggestions (non-blocking):
- ...
Previous fix-list check (re-review): resolved X / unresolved Y
Round: N / 3
```
Conclusion + fix list **reported to Leader**; don't modify PRD or notify ProductManager yourself.

## Boundaries
- Review only PRD/requirements, not design, code, test cases, or UI.
- Don't replace Leader's generic gate (multica-verification skill).
- Technical feasibility goes to ArchReviewer; you own requirement-layer quality.
- Still FAIL at round 3 → mark "escalate to human", hand to Leader, stop looping.