key-abstractions

Extract named domain units and grounding terms into a structured key-abstractions.md file.

1|Updated Feb 26, 2026
One-click install
npx skills add https://github.com/agilebydesign/agilebydesign-skills --skill key-abstractions
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: key-abstractions
Source: https://github.com/agilebydesign/agilebydesign-skills/tree/main/agents/abd-ooad/skills/key-abstractions
Command: npx skills add https://github.com/agilebydesign/agilebydesign-skills --skill key-abstractions

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Identify and extract named domain units (Key Abstractions) from source text to provide auditable, reusable vocabularies for downstream modeling and classification.

Core Features & Use Cases

  • Detect and name domain units as discrete abstractions with source-grounded intents.
  • Capture a concise Core terms list and a free-form Shape hint for each abstraction.
  • Output a structured key-abstractions.md that can be consumed by distillation and downstream design passes.

Quick Start

Run the key-abstractions workflow on your module-partitioning file if present, or on the raw corpus to generate key-abstractions.md.

Frequently Asked Questions about key-abstractions

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

FAQPage Schema
How do I extract domain abstractions from unstructured source text?

You extract domain abstractions by applying an analysis workflow to partitioned modules or unpartitioned corpora, which identifies named domain units and grounding terms to produce a structured key-abstractions.md file.

What is domain modeling distillation and when do I need it?

Domain modeling distillation is the process of identifying discrete domain units and their intents from source text. You need it to build auditable, reusable vocabularies that support downstream classification and design passes.

How do I document core terms and intents for requirements engineering?

Document core terms and intents by analyzing source text to capture concise term lists, free-form shape hints, and source-grounded intents for each named domain unit, ensuring every abstraction includes verbatim source extracts.

Can I process unpartitioned text corpora or do I need pre-partitioned modules?

You can process both unpartitioned corpora and partitioned modules. The workflow slices multi-subject passages into partials and documents unallocated slices, ensuring every key abstraction is grounded with source extracts regardless of input structure.

How does text analysis handle multi-subject passages when identifying key abstractions?

Text analysis handles multi-subject passages by slicing them into partials to isolate discrete domain units. When necessary, it documents unallocated slices to maintain an auditable link between extracted abstractions and verbatim source text.

What is the best way to structure knowledge representation for downstream design passes?

The best way to structure knowledge representation for downstream design is generating a key-abstractions.md file containing per-abstraction intents, core terms, shape hints, and verbatim extracts to provide a reusable, auditable vocabulary.