spec-requirements

Convert natural language requests into EARS-style requirements documents.

Updated Apr 21, 2026
One-click install
npx skills add https://github.com/LeeHusung/harness --skill spec-requirements-leehusung
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spec-requirements
Source: https://github.com/LeeHusung/harness/tree/main/.claude/skills/spec-requirements
Command: npx skills add https://github.com/LeeHusung/harness --skill spec-requirements-leehusung

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill converts a user's natural language request into the EARS-style requirements format, producing a structured spec that clearly defines outcomes, scope, constraints, decisions, tasks, and verification criteria.

Core Features & Use Cases

  • Translates natural language prompts into a formal 6-section requirements document following the spec guidelines.
  • Generates a complete .claude/specs/<YYYY-MM-DD>-<slug>/requirements.md ready for review and implementation.
  • Serves as a guardrail to ensure testable, unambiguous feature definitions for new work or substantial changes.

Quick Start

Provide the user request; the skill will generate a complete requirements.md document under .claude/specs with the appropriate date and slug.

Frequently Asked Questions about spec-requirements

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

FAQPage Schema
How do I convert natural language requests into formal product requirements?

To convert natural language requests into formal product requirements, this skill applies the EARS notation to generate a structured six-section requirements document. It guides users to define outcomes, scope, constraints, decisions, tasks, and verification criteria.

What is the EARS syntax for writing unambiguous specification requirements?

EARS is a structured syntax used to write unambiguous specification requirements. This skill automatically translates natural language feature requests into EARS-style statements to ensure testable and clear product definitions.

How do I generate a requirements.md file for a new feature request?

You generate a requirements.md file by providing a natural language prompt to the skill. It produces a complete spec file under the .claude/specs/<YYYY-MM-DD>-<slug>/ directory, ready for review and implementation.

When do I need to create a formal specification for a product change?

You need a formal specification for new features or substantial product changes. Creating an EARS-style spec ensures that the work has defined scope, constraints, and verification criteria before implementation begins.

Can I use natural language to define verification criteria for software specs?

Yes, you can use natural language to define verification criteria for software specs. The skill processes your plain text input and outputs deterministic, testable criteria within the formal EARS requirements document.