Build disposable prototypes to validate uncertain ideas with evidence-backed verdicts.
npx skills add https://github.com/MarbleSodas/Mavis --skill spike-marblesodas
Please help me install this Agent Skill. Skill: spike Source: https://github.com/MarbleSodas/Mavis/tree/main/skills/software-development/spike Command: npx skills add https://github.com/MarbleSodas/Mavis --skill spike-marblesodas
It helps you reduce the risk of building the wrong thing by turning uncertain ideas into quick, testable evidence before committing to production work.
Use the spike skill to validate an approach by saying: "Spike whether X is feasible for my use case and give me a VALIDATED/PARTIAL/INVALIDATED verdict with evidence."
High-intent search queries and answers about installing and using this skill.
Idea validation through throwaway experimentation involves building minimal interactable prototypes to test uncertain approaches and gather evidence. It reduces the risk of building the wrong thing by producing clear VALIDATED or INVALIDATED verdicts before committing to production work.
The best way to compare technical approaches is running parallel variants for head-to-head decisions. By building minimal prototypes for each approach, you can iterate based on observable evidence and document a clear recommendation for which path to choose for the real build.
Run a feasibility spike by breaking your idea into testable Given/When/Then questions, performing minimal research to pick an approach, and building an interactable prototype. You then document a per-spike README with a VALIDATED, PARTIAL, or INVALIDATED verdict backed by evidence.
Yes, you can use throwaway prototypes for head-to-head comparison by running parallel variants of your experiment. This allows you to test multiple solutions simultaneously, iterate based on the evidence each prototype produces, and confidently choose the best approach for your production build.
You should not use feasibility spikes when unknowns can be resolved through documentation review alone. Spikes are designed for early planning scenarios that require hands-on, interactable prototypes to generate observable evidence rather than relying on existing research or theoretical analysis.
jordanhubbardDecompose ideas into disposable experiments to validate feasibility before building.
JamesFincherDecompose proposals into feasibility questions and run disposable implementation spikes.
JuanMS20Decompose product ideas into throwaway spikes with verdicts and recommendations.
XWIlluDeluEvaluate scientific claims by critiquing methodology, bias, and statistical validity.
