spike

Decompose ideas into testable spike questions and build disposable prototypes.

Updated May 5, 2026
One-click install
npx skills add https://github.com/Z43L/zeus-agent --skill spike-z43l
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spike
Source: https://github.com/Z43L/zeus-agent/tree/main/skills/software-development/spike
Command: npx skills add https://github.com/Z43L/zeus-agent --skill spike-z43l

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) and references (resource) and scripts (resource) components.

What problem does it solve?

This Skill helps you de-risk an idea by running a lightweight, disposable feasibility experiment before committing engineering time.

Core Features & Use Cases

  • Feasibility decomposition: Break a proposal into 2–5 testable questions and run them as separate spikes.
  • Research-then-build loop: Do just enough investigation to choose an approach, then build an observable prototype.
  • Evidence-based verdicts: Produce a clear VALIDATED | PARTIAL | INVALIDATED outcome with what worked, what didn’t, surprises, and next-build recommendations.
  • Comparison-ready: Run competing approaches side-by-side (e.g., pdfjs vs camelot) and crown a winner with a head-to-head table.

Quick Start

Tell the agent: "Spike whether approach X is feasible for my project and give me a VALIDATED/PARTIAL/INVALIDATED verdict with a recommendation."

Frequently Asked Questions about spike

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

FAQPage Schema
How do I validate the technical feasibility of an idea before committing engineering time?

You validate feasibility by decomposing an idea into 2-5 testable spike questions and building a disposable prototype to gather observable evidence. This process de-risks your proposal by producing a clear VALIDATED, PARTIAL, or INVALIDATED outcome with next-build recommendations.

What is the engineering workflow for running a spike to test integration viability?

The workflow follows a decompose, research, build, and verdict loop. You first break down the integration unknowns into specific questions, do just enough research to choose an approach, build an observable prototype, and document the evidence to reach a final verdict.

How do I compare competing approaches side-by-side during a proof of concept?

You can compare competing approaches by running them as separate spikes with observable prototypes, then producing a head-to-head comparison table. This method evaluates the performance and behavior of each option side-by-side to crown a winning approach.

When do I need to build a disposable prototype for an MVP?

You need a disposable prototype when your MVP has unknowns requiring observable evidence, such as performance metrics, extraction behavior, or integration viability. It allows you to test specific spike questions and gather proof of concept data without committing to full implementation.

What evidence do I need to document when testing a spike question?

Per-spike documentation requires recording observable evidence from the prototype, detailing what worked, what did not work, and any surprises encountered. This evidence directly supports the final VALIDATED, PARTIAL, or INVALIDATED verdict and next-build recommendations.