spike

Validate feasibility by decomposing the idea into 2–5 independent Given/When/Then questions and building minimal disposable prototypes that produce real observable outputs. This method yields a structured VALIDATED, PARTIAL, or INVALIDATED verdict to guide your next steps.

A spike is a disposable prototype used to de-risk proposed builds when research alone is insufficient. Use a spike for multi-approach exploration, unknown integration risks, and uncertainty-heavy features to generate concrete feasibility evidence before committing to full implementation.

De-risk integration issues by building the smallest interactive artifact possible, such as a CLI, small HTML page, or tiny server, to test specific feasibility questions. This evidence-producing approach replaces guesswork with observable results and a structured recommendation.

You can test multiple approaches by framing each as an independent, observable feasibility question with explicit Given/When/Then criteria. The process builds minimal targeted artifacts per question, producing comparable verdicts to recommend what to do for the real build.

Feasibility validation requires only minimal targeted research before building a spike. The focus shifts quickly to creating an observable artifact that produces real outputs, ensuring you gather concrete evidence efficiently rather than relying solely on preliminary research.

Use disposable prototypes because research alone is often insufficient for uncertainty-heavy features. Prototypes produce real observable outputs and evidence rather than assumptions, yielding a structured verdict that directly recommends the next action for your real build.