How does the skill avoid asking obvious questions?

It performs a silent Phase 1 'Context Gathering' where it explores your codebase's existing architecture, models, and naming conventions to ensure the interview is informed and relevant.

How many questions will the AI ask during the interview?

The skill calibrates based on scope: Small changes trigger 2-4 questions, Medium features 4-8 questions, and Large architectural changes can involve 8-15 questions.

Where does the skill save the final implementation plan?

By default, the plan is written to a Markdown file in the docs/plans/ directory, or the project root if a documentation folder does not exist.

What is the primary purpose of the interview skill?

The interview skill is designed to help users think through a feature, clarify technical requirements, and produce a structured specification file before any coding begins.

Can I use this skill to generate code directly?

No, this skill is specifically for the planning and design phase. It produces a written specification that serves as a roadmap for subsequent coding tasks.

Requirement Interviewer

Name: Requirement Interviewer
Author: atc-net

byatc-net

•

Productividad y Flujo de Trabajo

Conducts structured interviews to transform vague feature ideas into concrete technical specifications and implementation plans.

The interview skill empowers Claude to act as a lead architect and product manager by guiding you through a systematic requirement-gathering process. Before writing any code, the skill silently analyzes your existing project architecture to ensure informed questioning. It then leads a calibrated interview—ranging from discovery to implementation details—to identify edge cases, define scope, and establish technical patterns. The process culminates in a professional Markdown specification file, ensuring your development work is grounded in a solid, documented plan.

Características Principales

011 GitHub stars

02Scope-based calibration that adjusts the number of questions to the project size

03Comprehensive coverage of happy paths, edge cases, and data models

04One-at-a-time focused questioning to prevent user fatigue

05Automatic generation of technical plans in Markdown format

06Silent codebase exploration to understand existing patterns and domain concepts

Casos de Uso

01Mapping out complex edge cases and error handling strategies before implementation

02Collaborating with an AI agent to clarify vague business requirements into actionable tasks

03Defining the technical architecture for a new subsystem or feature extension

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add atc-net/atc-agentic-toolkit interview

For use in Claude.ai and ChatGPT

Características Principales

011 GitHub stars

02Scope-based calibration that adjusts the number of questions to the project size

03Comprehensive coverage of happy paths, edge cases, and data models

04One-at-a-time focused questioning to prevent user fatigue

05Automatic generation of technical plans in Markdown format

06Silent codebase exploration to understand existing patterns and domain concepts

Casos de Uso

01Mapping out complex edge cases and error handling strategies before implementation

02Collaborating with an AI agent to clarify vague business requirements into actionable tasks

03Defining the technical architecture for a new subsystem or feature extension