walking-skeleton

Captures the thinnest end-to-end architectural slice and hands the plan to /to-issues.

Updated Jun 2, 2026
One-click install
npx skills add https://github.com/MohamedSayedK/omnigrid --skill walking-skeleton-mohamedsayedk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: walking-skeleton
Source: https://github.com/MohamedSayedK/omnigrid/tree/main/assets/core/skills/walking-skeleton
Command: npx skills add https://github.com/MohamedSayedK/omnigrid --skill walking-skeleton-mohamedsayedk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? New features often stall because integration risk is discovered late. This Skill forces you to define the thinnest end-to-end slice of a feature — one trivial happy path wired through every architectural layer — so integration and architecture are proven before real business logic is built. ## Core Features & Use Cases - Structured skeleton capture: Collects layers crossed, the thinnest happy path, proof-of-life, and an explicit in/out scope boundary through one-question-at-a-time conversation. - Taxonomy guidance: Distinguishes the walking skeleton (kept, production-shaped) from throwaway spikes and prototypes, redirecting when the wrong technique is chosen. - Handoff to /to-issues: Produces a confirmed structured plan with acceptance criteria and passes it to /to-issues for ticket filing. - Use Case: Before building a new orders service, use this Skill to define a slice where POST /orders with a hardcoded line item returns 201 and writes one row, proving UI, API, domain, and datastore layers are wired end to end. ## Quick Start Ask the assistant to run /walking-skeleton for the feature or service you want to prove end to end, then answer its questions one at a time.

Frequently Asked Questions about walking-skeleton

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

FAQPage Schema
What is a walking skeleton in software development?▼

A walking skeleton is the thinnest end-to-end slice of a feature wired through every architectural layer with no business logic beyond one trivial happy path. Unlike a spike or prototype, it is kept production-shaped code that the real product is built on top of.

How do I define a walking skeleton for a new feature?▼

Run /walking-skeleton with the feature name, then answer its questions one at a time: list every layer crossed, describe the single thin happy path, state the proof-of-life check, and draw the in/out scope boundary. The Skill confirms the plan and hands it to /to-issues.

Walking skeleton vs spike vs prototype: what is the difference?▼

A spike answers a technical question and is thrown away; a prototype explores UX direction and is thrown away; a walking skeleton proves the whole architecture is wired end to end and is kept as the production spine. The Skill redirects you to /spike or /prototype when those fit better.

When should I not use a walking skeleton?▼

Skip it when you only need a throwaway technical answer (use /spike) or a throwaway UX direction (use /prototype). It is also premature if you cannot enumerate the architectural layers, since the skeleton exists to prove integration across all of them.

Does the walking-skeleton skill create tickets directly?▼

No. The Skill only captures and confirms the structured skeleton plan, then hands it to /to-issues, which is the single ticket-cutter. Ticket creation is deliberately out of scope for this Skill.