spike

Plan and execute throwaway experiments to validate ideas before full builds.

Updated Jun 11, 2026
One-click install
npx skills add https://github.com/LamseyahElias/jarvis-cloud-v2 --skill spike-lamseyahelias
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spike
Source: https://github.com/LamseyahElias/jarvis-cloud-v2/tree/main/hermes-agent/skills/software-development/spike
Command: npx skills add https://github.com/LamseyahElias/jarvis-cloud-v2 --skill spike-lamseyahelias

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Throwaway experiments to validate ideas before committing to a full build.

Core Features & Use Cases

  • Decompose ideas into 2-5 independent feasibility questions and frame them as spikes.
  • Align, Research, Build, and Verdict steps to iterate and surface unknowns.
  • Use parallel spikes via delegate_task to compare approaches and pick the best path.

Quick Start

Create a new spikes/NNN-descriptive-name directory with a README and a minimal runnable artifact to demonstrate the spike output.

Frequently Asked Questions about spike

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is a spike in software engineering and when should I use it?

A spike is a throwaway experiment to validate ideas before committing to a full build. You should use a spike for early feasibility studies, comparing approaches, and surfacing unknowns with observable outputs.

How do I plan and execute a feasibility experiment for a new feature?

To plan a feasibility experiment, decompose ideas into 2-5 independent questions and frame them as spikes. Execute them by following the Align, Research, Build, and Verdict steps to iterate and surface unknowns.

Can I run parallel spikes to compare different technical approaches?

Yes, you can run parallel spikes via delegate_task to compare approaches and pick the best path. This allows you to evaluate multiple independent feasibility questions simultaneously before a full build.

What is the best way to structure a spike directory?

The best way to structure a spike directory is to create a new spikes/NNN-descriptive-name directory. This directory should contain a README and a minimal runnable artifact to demonstrate the spike output.

How do I decide if a spike validated or invalidated an idea?

To decide if a spike validated an idea, use a clear verdict process that outputs VALIDATED, PARTIAL, or INVALIDATED. This structured verdict guides future development based on the experiment's observable results.