plan-ceo-review

Files (1)

• SKILL.md

SKILL.md (first lines)

# CEO Review

Review a plan from a founder or product-lead perspective. Re-evaluate the user problem, target user, wedge, and value of the proposed scope rather than treating the plan as fixed.

Use this skill when the user asks to think bigger, challenge or reduce scope, review product direction, or judge whether a plan is focused enough before implementation.

If missing product context would materially change the review and cannot be inferred from the supplied plan or surrounding context, ask the smallest question needed. Otherwise state reasonable assumptions and complete the review.

## Review lens

Choose the lens that best matches the user's intent; do not require a separate confirmation:
- **Scope expansion:** identify a meaningfully stronger product direction when the current plan is too timid.
- **Selective expansion:** preserve the core plan and add only high-leverage capabilities.
- **Hold scope:** protect focus, schedule, and the existing wedge.
- **Scope reduction:** remove speculative or low-leverage work while preserving real user value.

## Review process

1. Restate the real user problem, target user, proposed solution, and apparent boundaries.
2. Challenge the premise: test whether the problem is real, the user is concrete, the outcome is visible, and the wedge is sharp.
3. Evaluate whether scope is too small to matter, too broad to focus, missing a key capability, or prematurely optimized.
4. Explain the material tradeoffs and give a concrete product direction consistent with the chosen lens.
5. Identify any decision that must remain with the user only when different choices would materially change the product direction.

## Output

Use a concise structure that covers:
- **Premise:** problem, beneficiary, and key assumption.
- **Scope assessment:** what is strong, weak, missing, or unnecessary.
- **Risks:** material product, adoption, and focus risks.
- **Direction:** what to add, keep, remove, or reframe and why.
- **Open decision:** include only when user input is actually required before the next step.
- **Next step:** the most useful action before implementation.

Tie findings to user value and observable outcomes. Avoid generic feature requests, implementation trivia without product impact, or scope changes without a clear reason.

This skill works from the plan and available local context; it does not depend on repository-specific status systems.

From minfengyuan/harness-knowledge-base, MIT.

More from minfengyuan/harness-knowledge-base