skills.rest
tapway
tapwayOFFICIAL
Explore all skills by this author →
WebsiteGitHub

gbrain-media-ingest

Extract transcripts, OCR text, and descriptions from media files into gbrain.

1|Updated Jun 21, 2026
Search Tags:#ocr#knowledge base#transcription#whisper#entity extraction#gbrain#media ingestion
One-click install
npx skills add https://github.com/tapway/shogun-os --skill gbrain-media-ingest-tapway
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gbrain-media-ingest
Source: https://github.com/tapway/shogun-os/tree/main/skills/gbrain/gbrain-media-ingest
Command: npx skills add https://github.com/tapway/shogun-os --skill gbrain-media-ingest-tapway

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Media content like videos, voice notes, PDFs, and screenshots cannot enter a knowledge base until it is transcribed, OCR-extracted, or visually described. This Skill converts media into structured, entity-linked brain pages so nothing gets lost. ## Core Features & Use Cases - Multi-format extraction pipelines: Transcribe YouTube videos and audio via Whisper, OCR PDFs with pymupdf or marker-pdf, and describe screenshots with vision analysis. - Entity extraction and cross-linking: Identify people, companies, projects, and dates, then link new pages to existing brain entries and add timeline entries. - Verbatim preservation: Voice notes and direct quotes are stored with exact phrasing, while long-form content is summarized. - Use Case: A user sends a voice note saying a client wants to renew in Q3. The Skill transcribes it verbatim, links it to the client's company page, and adds a timeline entry for the renewal signal. ## Quick Start Ingest this voice note into gbrain, transcribe it exactly, extract the entities, and link it to the matching company page.

Frequently Asked Questions about gbrain-media-ingest

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

FAQPage Schema
How do I ingest a voice note into a knowledge base?▼

Transcribe the audio with Whisper or an equivalent API, preserving the exact phrasing without paraphrasing. Then extract entities like people and companies, search the brain for matches, write the page under ideas/ or references/media/, and cross-link it.

How to extract text from PDF documents for a knowledge base?▼

Use OCR tools such as pymupdf or marker-pdf to extract text from PDFs, then summarize key points and extract entities. File the result under references/<topic>/ or projects/<project>/ and link it to related brain pages.

Can screenshots and images be stored in a searchable knowledge base?▼

Yes. Use vision analysis to describe the visual content and extract any visible text, then summarize and extract entities. The resulting page is filed under references/media/ or a topic path with cross-links to related entities.

When should I not use media ingestion for content?▼

Do not use it for plain text or markdown files, which go through standard text ingestion, or for web articles without media. It is also wrong for paraphrasing voice notes, since verbatim phrasing must always be preserved.

Why is exact phrasing important when transcribing voice notes?▼

Voice notes and direct quotes lose meaning and attribution when paraphrased, so the pipeline stores the verbatim transcript. Summarization is reserved for long-form video and audio extracts, not short direct statements.

Related Skills

Explore More Product & Management Skills→
enclawedenclawed
2

qqbot-media

Route QQBot media messages to correct attachment types from local paths or URLs.

Community
Intermediate
yunusgungoryunusgungor

meeting-ingestion

Parse meeting transcripts into brain pages with attendee enrichment and timeline merge.

Community
Advanced
edwifiguyedwifiguy
1

idea-ingest

Ingest and analyze shared links, articles, tweets, and ideas into the brain.

Community
Advanced
jaminitachijaminitachi
1

memory-mgr

Manage persistent cross-session memory with a knowledge graph and full-text search.

Community
Advanced