| name | plan-ceo-review |
| description | Structured CEO-level review of any PRD or feature plan. Four scope modes. Challenges assumptions, finds the 10-star product hiding inside the request. |
| used_by | ["product-manager","ceo-thinking-partner"] |
CEO Plan Review Skill
When product-manager produces a PRD, CEO runs this review before approving.
Four Scope Modes (CEO picks one)
- EXPANSION — "This is too small. What's the 10-star version?"
- SELECTIVE EXPANSION — "Keep core scope, but find one area to go deeper"
- HOLD SCOPE — "Scope is right. Stress-test the details."
- REDUCTION — "This is too big. What's the narrowest wedge that validates the hypothesis?"
10-Section Review Checklist
For every PRD, evaluate:
- Problem Statement — Is the pain real? Evidence from TRACEABILITY_MATRIX.md? Or assumption?
- User Segments — Does this serve [DOMAIN_ACTOR_1]s? [DOMAIN_ACTOR_2]s? Both sides of the marketplace?
- Scope — What's in? What's explicitly out? Is the out-list honest?
- Success Metrics — How do we know this worked? Measurable within 2 weeks of launch?
- Dependencies — What must exist first? Schema? API? Auth changes? Flag Tier 3 gates.
- Risks — What kills this? Compliance risk? Technical risk? Market risk?
- Alternatives Considered — Did PM evaluate at least 2 other approaches? Why was this one chosen?
- Timeline — Is this a 1-sprint or multi-sprint feature? Sprint allocation realistic?
- Marketplace Impact — Using the two-sided marketplace rule: does this serve both sides?
- Recommendation — APPROVE / REVISE / REJECT with specific reasoning
Output Format
Write review to docs/plan-reviews/[feature]-ceo-review-[date].md
Post to Slack CEO: "CEO PLAN REVIEW: [feature] — [APPROVE/REVISE/REJECT] — [1-line reason]"
If REVISE: product-manager re-works and re-submits.
If REJECT: feature goes to Parked Items with documented reasoning.
