What audio formats are supported for transcription?

The skill provides workflows for various services including Otter.ai, Rev.com, and local Whisper models, helping you convert most common audio formats into markdown.

Can I use this for batch processing?

Yes, the skill includes dedicated scripts like process-audio.js to batch process multiple recordings and documents into your project structure.

How does it handle complex PDF elements like tables?

In Tier 1 mode, it uses VLM-powered parsing (MinerU) to extract tables and formulas with high accuracy, maintaining their structure for later analysis.

Where is the converted data stored?

Converted files are organized into a standardized 'stage1-foundation/manual-codes/' directory with an accompanying JSON inventory for tracking.

Does this skill require a paid API key?

While Tier 1 features use the MinerU API for high-accuracy PDF parsing, the skill includes Tier 2 manual workflows that work without any additional API keys.

Interview Ingest & Data Prep

Name: Interview Ingest & Data Prep
Author: linxule

bylinxule

•

データサイエンスとML

Converts audio recordings, PDFs, and diverse document formats into structured markdown for qualitative research and AI analysis.

The interview-ingest skill streamlines the foundational phase of qualitative research by transforming raw data sources—such as interview audio and complex PDF documents—into AI-ready markdown files. It supports a two-tier capability system, utilizing advanced VLM-powered parsing through MinerU for high-accuracy PDF extraction or providing robust manual workflows for researchers without specialized API keys. By automating the conversion process and maintaining a structured data inventory, this skill ensures that researchers can quickly transition from data collection to synthesis while maintaining a clear audit trail between original sources and transcribed content.

主な機能

01Multi-platform audio transcription support for Otter.ai, Rev, and Whisper

02Structured markdown output optimized for AI-assisted qualitative coding

03Dual-tier PDF parsing using MinerU VLM or manual OCR workflows

04Batch processing scripts for managing large sets of interview recordings

05Automatic data inventory management with JSON tracking for audit trails

062 GitHub stars

ユースケース

01Standardizing diverse data sources into a unified project repository for stage-based research

02Transcribing academic interviews into structured markdown for thematic analysis

03Extracting data from complex, multi-column PDF research papers and tables

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add linxule/interpretive-orchestration interview-ingest

For use in Claude.ai and ChatGPT

Download Skill