What are LLMREI patterns?

LLMREI refers to Large Language Model Requirements Elicitation Interview research-backed patterns designed to maximize the accuracy and completeness of requirements captured by AI agents.

How does Guided mode differ from Full-Auto?

Guided mode asks for your approval before posing each question to a stakeholder, while Full-Auto conducts the entire interview autonomously and provides a summary at the end.

What happens if I don't have a stakeholder available?

You can use the simulation feature to have Claude spawn a specific persona (like a Security Officer or End User) to explore requirements and potential constraints from that perspective.

Where are the interview results saved?

Results are saved as structured YAML files in the .requirements/{domain}/interviews/ directory of your project for easy versioning and reference.

Can I use this with real stakeholders?

Yes, in the default mode, Claude acts as the interviewer and communicates directly with the person in the chat to extract project details and clarify needs.

Requirements Interviewer

Name: Requirements Interviewer
Author: melodic-software

bymelodic-software

•

コラボレーションツール

Conducts structured, research-backed requirements elicitation interviews with stakeholders to capture project needs accurately.

This skill automates the discovery phase of software development by leading stakeholders through research-backed LLMREI interview patterns. Whether interviewing a real Product Manager or simulating a technical persona, the skill systematically gathers functional requirements, non-functional constraints, and business logic. It helps teams bridge the gap between initial ideas and formal specifications by identifying themes and gaps automatically, saving results in a structured format for easy integration into the development lifecycle.

主な機能

01Automatic identification and categorization of functional and non-functional requirements

02Structured output generation in YAML for seamless documentation and project integration

0338 GitHub stars

04Flexible autonomy modes including guided, semi-auto, and full-auto

05Research-backed LLMREI interview patterns for high-quality requirement capture

06Stakeholder simulation for exploring diverse perspectives and testing assumptions

ユースケース

01Facilitating discovery sessions with Product Managers to define new feature scopes

02Gathering technical constraints from Lead Engineers for architectural planning

03Simulating end-user feedback to identify potential usability gaps early in development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add melodic-software/claude-code-plugins interview

For use in Claude.ai and ChatGPT

Download Skill