ck-prototypelisted

# CK Prototype Use a disposable artifact to make a decision cheaply. The output is a supported direction and the experiment that informed it. Production implementation follows the user's chosen scope. ## 1. Name the decision Identify the uncertainty the prototype will resolve: layout, interaction, behavior, timing, interface shape, or an analysis method's suitability. State the input conditions and observations that would distinguish the live alternatives. Use the user's settled goals and constraints. If the task is already a straightforward implementation request with no decision to explore, proceed with that request instead of manufacturing an experiment. ## 2. Explore plausible directions When the design space is open, inspect relevant existing work and references. Offer useful variations beyond the first idea when they address the decision. If a consequential user preference is missing, clarify it; otherwise build from the supplied direction and stated assumptions. Choose enough variants to expose the actual tradeoff. A focused experiment can also answer one hypothesis without a comparison set. ## 3. Build the smallest useful experiment Work in a clearly named scratch directory, separate from production source and original data. Use the project's available runtime and the lightest implementation that can demonstrate the behavior. Favor speed and observability over reusable abstractions or production polish. - **Visual or interaction questions:** make variants ea