spike

Decompose feasibility questions into throwaway prototypes and return a structured verdict.

Updated May 10, 2026
One-click install
npx skills add https://github.com/Mateus2411/Hermes-PersonalBot --skill spike-mateus2411
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spike
Source: https://github.com/Mateus2411/Hermes-PersonalBot/tree/main/skills/software-development/spike
Command: npx skills add https://github.com/Mateus2411/Hermes-PersonalBot --skill spike-mateus2411

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you de-risk an idea by turning uncertainty into quick, observable feasibility results before investing time in a full build.

Core Features & Use Cases

  • Decompose feasibility into testable spike questions so each risk has an explicit Given/When/Then outcome.
  • Research just enough to choose an approach by comparing options, consulting docs, and checking what tools are available.
  • Build throwaway prototypes with evidence (CLI output, minimal HTML, tiny server endpoint, or focused tests) to avoid false “it works” conclusions.
  • Produce a clear verdict and recommendation (VALIDATED | PARTIAL | INVALIDATED) with what worked, what didn’t, surprises, and what to do next.
  • Run comparison spikes in parallel (e.g., two extraction approaches) and finish with a head-to-head winner.

Quick Start

Tell the AI: "Spike whether my approach X is feasible for Y, and return a verdict with evidence and a recommendation for the real build."

Frequently Asked Questions about spike

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

FAQPage Schema
How do I validate a proof-of-concept before committing to full development?

To validate a proof-of-concept, you run a spike: a throwaway feasibility experiment that tests a specific idea with minimal research and a targeted prototype, returning a structured verdict before you build production code.

What is a spike in engineering planning and when should I use it?

A spike in engineering planning is a quick, isolated experiment used to decompose uncertainty into testable Given/When/Then questions. Use it to compare approaches or test feasibility before investing time in a full production build.

How do I compare technical approaches and choose the best one for my project?

You compare technical approaches by running parallel spikes: building isolated throwaway prototypes for each option, testing edge cases, and returning a head-to-head winner with a clear recommendation for the real build.

Can I test the feasibility of an idea without writing production code?

Yes, you can test feasibility without production code by building throwaway prototypes that produce observable evidence like CLI output, minimal HTML, tiny server endpoints, or focused tests to avoid false positive conclusions.

What is the best way to de-risk a technical idea before building it?

The best way to de-risk a technical idea is to create a spike plan that researches just enough to choose an approach, builds a minimal prototype, and returns a VALIDATED, PARTIAL, or INVALIDATED verdict with recommendations.

Why do proof-of-concept tests sometimes return false positive results?

Proof-of-concept tests return false positives when they lack observable evidence. Building throwaway prototypes with CLI output or focused tests ensures you avoid false 'it works' conclusions and accurately invalidate unfeasible ideas.