skills.rest
ngochuy13
ngochuy13COMMUNITY
Explore all skills by this author →
GitHub

media-ingest

Convert heterogeneous media inputs into structured brain pages with entity backlinks.

Updated Apr 23, 2026
Search Tags:#OCR#entity extraction#YouTube transcripts#media ingest#backlink propagation#GitHub summarization
One-click install
npx skills add https://github.com/ngochuy13/intern-dev --skill media-ingest-ngochuy13
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: media-ingest
Source: https://github.com/ngochuy13/intern-dev/tree/main/skills/media-ingest
Command: npx skills add https://github.com/ngochuy13/intern-dev --skill media-ingest-ngochuy13

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns messy media inputs (video, audio, PDFs, books, screenshots, and GitHub repos) into coherent brain pages with summaries, highlights, and linked entities instead of leaving you with transcripts or isolated notes.

Core Features & Use Cases

  • Multi-format ingestion with analysis: Extracts text and entities from varied media types (including OCR and repo summarization) and writes a real brain page with a structured summary, not a transcript dump.
  • Entity extraction + backlink propagation: Detects every person and company mentioned, ensures those entity pages exist/enriches as needed, and creates backlinks and timeline entries.
  • Provenance-first raw preservation: Uploads the original source file for traceability before creating the brain page, then syncs the index.

Quick Start

Ingest the provided file or link by asking: ingest this media into the brain and return the number of entities found plus which pages were updated.

Frequently Asked Questions about media-ingest

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

FAQPage Schema
How do I extract text and entities from YouTube video transcripts?▼

During media ingest, entity extraction detects people and companies from video transcripts, creates their respective entity pages, and propagates backlinks with timeline entries.

Can I extract text from PDFs and screenshots using OCR?▼

The ingestion process applies OCR to screenshots and PDF book text, preserving the raw source file for provenance while generating a summarized brain page.

What is the best way to summarize a GitHub repository architecture?▼

Ingesting a GitHub repository generates a structured brain page with architectural summaries, highlights, and cross-linked entities, avoiding isolated transcript or code dumps.

How does backlink propagation work for extracted entities?▼

During ingestion, backlink propagation ensures all detected person and company entity pages exist, enriches them, and creates reciprocal backlinks with timeline entries.

Does media ingestion preserve the original source file?▼

Yes, provenance-first raw preservation uploads the original source file for traceability before generating the structured brain page and synchronizing the final index.

Related Skills

Explore More Content & Communication Skills→
ltdrewltdrew
1

media-ingest

Ingest video, audio, PDF, book, screenshot, and GitHub repo content into brain pages with entity extraction and backlink propagation.

Community
Advanced
ngochuy13ngochuy13

meeting-ingestion

Ingest meeting transcripts into a knowledge graph with enriched attendees and entity links.

Community
Advanced
SigmacodeatSigmacodeat

media-ingest

Convert media content into structured knowledge pages with entities and relationships.

Community
Advanced
NinatuziNinatuzi

media-ingest

Ingest videos, audio, PDFs, and GitHub repositories into a structured knowledge graph.

Community
Advanced