Prototype Plan

---
name: prototype-plan
description: Plan a prototype around a specific design or product uncertainty, selecting the minimum fidelity, scope, participants, scenarios, evidence, and success criteria needed to inform a decision.
---
# Prototype Plan

Use when a product interaction, flow, information model, or design direction contains uncertainty that should be tested before implementation.

## Procedure
1. State the decision the prototype must inform and the assumption at risk. Avoid prototyping an entire feature when only one interaction or mental-model question is uncertain.
2. Choose fidelity based on the question: sketch/wireframe for structure, clickable prototype for navigation and comprehension, coded prototype for browser/device behavior or technical interaction constraints, or another form that can actually expose the risk.
3. Define the smallest scenario and states participants need to experience. Include enough realistic content/data to make the decision meaningful without polishing unrelated surfaces.
4. Decide who should evaluate the prototype and why. For usability questions, route participant/research design to `agency-user-researcher`; for technical feasibility, coordinate with the relevant engineer; for internal design critique, identify the decision owner.
5. Write representative tasks or prompts that reveal whether users understand and can complete the intended action without teaching them the solution.
6. Define evidence and decision thresholds in advance where practical: task completion, comprehension, observed confusion, preference with rationale, time/steps, qualitative patterns, or technical feasibility constraints relevant to the assumption.
7. Include accessibility considerations appropriate to the fidelity. A low-fidelity visual prototype may not prove assistive-technology behavior, so do not overclaim what it validates.
8. Plan capture: observations, screen/interaction recordings when consent and context allow, notes, participant quotes, failure points, or technical measurements. Preserve raw evidence separately from interpretation.
9. Define what happens after the prototype: proceed, revise, compare another approach, run formal user research, hand off to engineering, or stop the idea.

## Decision rules
- Prototype the riskiest decision, not the prettiest screen.
- Fidelity should be no higher than required to answer the question.
- A prototype demonstrates a possible experience; it is not automatically production specification or proof of demand.
- Do not treat stakeholder preference as user validation.
- When the uncertainty is product desirability rather than interaction usability, coordinate with Product Manager/User Researcher on the correct evidence method.

## Quality gate
The plan is ready when the prototype has one clear decision purpose, fidelity matches the uncertainty, scope is minimal but realistic, evaluation participants/scenarios and evidence are defined, limitations are explicit, and a concrete next decision follows from the result.