How does it handle uncertainty in the information gathered?

It uses epistemic confidence tracking to assign scores to findings, identifying whether gaps are due to lack of information (epistemic), inherent randomness (aleatory), or framework choices (model).

Can I choose different levels of rigor for the interview?

Yes, it supports three validation modes: Empathetic (focuses on rapport), Balanced (standard verification), and Rigorous (aggressive assumption challenging and devil's advocate probing).

What is the Research Interviewer skill?

It is a specialized capability for Claude that conducts structured, adaptive interviews to extract detailed and validated information for research, documentation, or development.

How does this skill ensure all topics are covered?

The skill uses a MECE (Mutually Exclusive, Collectively Exhaustive) decomposition framework to map out the topic space and ensure there are no overlapping categories or missing dimensions.

Research Interviewer

Name: Research Interviewer
Author: Agentient

byAgentient

0•

Learning & Documentation

Conducts systematic, high-fidelity knowledge elicitation through structured interviewing and epistemic confidence tracking.

Research Interviewer is a sophisticated tool designed to bridge the gap between vague ideas and structured documentation by conducting adaptive, multi-phase interviews. It employs MECE (Mutually Exclusive, Collectively Exhaustive) decomposition to ensure full coverage of a topic, tracks epistemic confidence to identify information gaps, and applies bias-protection techniques to ensure objective data collection. Whether you are eliciting complex software requirements, building a knowledge base for RAG systems, or defining problem statements, this skill ensures that every detail is captured, validated, and synthesized into a high-quality, actionable deliverable.

Key Features

01Adaptive questioning using eight distinct elicitation strategies for deep probing

020 GitHub stars

03MECE scope mapping to ensure comprehensive coverage of any topic

04Proactive bias protection and 'Steelmanning' to validate and strengthen findings

05Real-time epistemic confidence tracking to quantify the reliability of gathered information

06Multi-format output generation for problem statements, requirements, or knowledge bases

Use Cases

01Extracting domain expertise to build high-fidelity knowledge bases for RAG systems

02Gathering comprehensive software requirements from stakeholders before development begins

03Defining precise problem statements and research briefs for complex strategic projects

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add agentient/vibekit research-interviewer

For use in Claude.ai and ChatGPT

Download Skill