Build an AI-assisted PRD critique workflow

Description

Your PRD reviews are bottlenecked on senior PMs who read every draft. This builds an AI-assisted PRD critique workflow — checklist prompt, specific quality gates, human review for high-stakes items — so drafts improve before a human reads them and seniors focus on the hard problems.

You are designing an AI-assisted PRD critique workflow for {{team_name}}.

## Workflow steps
### Step 1 — PM drafts PRD using team template
### Step 2 — PM runs AI critique prompt
The critique prompt checks:
- Problem framing (is it a problem or a solution disguised as one?)
- Evidence (is there customer research backing the claim?)
- Success metrics (measurable, aligned to north star)
- Scope (is "not doing" list explicit?)
- Risks (named, with mitigation)
- Alternatives considered
- Stakeholder alignment
- Dependency mapping

### Step 3 — PM revises based on critique
### Step 4 — AI re-critiques
### Step 5 — Human review (senior PM) only on items AI flags uncertain OR is clearly out-of-scope

## Quality gates
- AI critique surfaces ≥3 specific gaps per first draft
- PM revisions address ≥80% of gaps before human review
- Human reviewer spends

Customize This Prompt