One-click install
npx skills add https://github.com/RCSnyder/lights-out-swe-plugin --skill expand-rcsnyder
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: expand
Source: https://github.com/RCSnyder/lights-out-swe-plugin/tree/main/skills/expand
Command: npx skills add https://github.com/RCSnyder/lights-out-swe-plugin --skill expand-rcsnyder

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns a vague “build this” idea into a structured, agent-ready scope that defines problem context, the smallest useful version, acceptance criteria, stack, deployment target, and the data model.

Core Features & Use Cases

  • Evidence-based scoping: Reads docs/input/ (preferring distilled- versions when available) to convert briefs and specs into measurable acceptance criteria and constraints.
  • Preferences-aligned planning: Uses preferences.md to select a stack and deployment target, and flags conflicts for resolution when user intent contradicts the configured preferences.
  • Agent-gated output: Produces scaffolding/scope.md with verifiable AC-* IDs and clear guardrails so later phases can validate progress.

Quick Start

Run the lo-swe init/expand flow with a one-liner describing your new project, and it will create scaffolding/scope.md containing acceptance criteria, stack, deployment target, and data model.

Frequently Asked Questions about expand

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

FAQPage Schema
How do I generate a project scope from a one-liner build request?

To generate a project scope, you provide a one-liner describing your new software project. The tool reads optional input briefs from docs/input/ and required constraints from preferences.md to create a structured scope document with acceptance criteria, stack, and deployment target.

How do I convert vague software ideas into measurable acceptance criteria for AI agents?

Converting vague ideas into measurable acceptance criteria involves expanding your one-liner request into a scope document with stable AC-* identifiers. It reads sourced documents and distilled briefs to define verifiable thresholds and clear guardrails for agent execution.

Does this project scoping tool work with my existing preferences.md and docs/input/ files?

Yes, project scoping works directly with your preferences.md to select stack and deployment target, and reads docs/input/ for briefs and specs. It flags conflicts for resolution when user intent contradicts configured preferences.

What is the best way to define the smallest useful version of a new software project?

Defining the smallest useful version requires deriving a coherent scope from user input and sourced documents. The tool generates scaffolding/scope.md containing problem context, data model, and verifiable acceptance criteria to establish a functional baseline.

Can I use a generated project scope document to validate AI coding agent progress?

Yes, you can use the generated scope document to validate agent progress. It produces scaffolding/scope.md with agent-gated output, stable AC-* IDs, and measurable thresholds so later development phases can verify completion against defined guardrails.

Why does project scope generation flag conflicts between user intent and configured preferences?

Project scope generation flags conflicts to ensure preferences-aligned planning when user intent contradicts configured preferences. This mechanism prevents architectural mismatches by requiring resolution before finalizing the stack and deployment target in the scope document.