mvp-scope-validator

MVP Scope Validator

The Law

AN MVP THAT CANNOT NAME ITS HYPOTHESIS IS NOT AN MVP.
"We need all these features to launch" is a bet on multiple unvalidated assumptions at once.
One hypothesis → minimum scope to test it → binary success metric IS a validated learning.

When to Use

• When defining a new product or feature for the first time
• When reviewing a PRD or feature list that feels too large
• When the team debates what to cut before a deadline
• When a feature list has grown and the original hypothesis is no longer clear

When NOT to Use

• For an existing product with live users — this validates new hypotheses, not iterative improvements
• When the idea is so vague that no hypothesis can be extracted — the skill will flag LOW_HYPOTHESIS and ask first
• For a single required bug fix or compliance task — there is no MVP for a fix; the scope is already defined
• When the goal is not learning (e.g., a required regulatory feature) — compliance requirements cannot be reduced by hypothesis framing

Inputs

Input	Required	Description
idea	required	Product or feature description — can be a sentence, a paragraph, or a feature list
constraints	optional	Time, budget, team size, or tech stack constraints that affect feasibility
hypothesis	optional	If the user already has a hypothesis, use it as the anchor; otherwise extract it from the idea

Outputs

Output	Description
mvp-definition	Core hypothesis, in-scope features (minimum to test hypothesis), out-of-scope items (explicitly removed), success metric
scope-reduction	List of removed items with a one-sentence reason per item

Steps

1. Read the idea. Try to extract a core hypothesis in the form: "We believe [target user] will [desired behaviour] because [reason]."
2. If no specific hypothesis can be extracted from the idea, set context-flag to LOW_HYPOTHESIS and ask 3 specific questions to help the user define one. Stop here until the hypothesis is clarified.
3. List every feature or capability mentioned in the idea.
4. For each feature, ask: "Is this required to test the hypothesis?" If not, it is out of scope for the MVP.
5. If the scope cannot be reduced at all (every feature is required to test the hypothesis), set context-flag to ALREADY_MINIMAL and explain why.
6. Write the in-scope feature set — the minimum to test the hypothesis.
7. Write the out-of-scope list: each item with a one-sentence reason why it is not needed for the hypothesis test.
8. Define the success metric as a binary statement: "The hypothesis is validated if [specific measurable outcome] within [timeframe]." The metric must be observable without interpretation.
9. Output the MVP definition and scope-reduction list.
10. Fill the Completion Statement.