youtube-ingestion-agent

Convert raw YouTube transcripts into validated study notes with controlled tagging and approval gates.

3|4|Updated Jul 20, 2026
One-click install
npx skills add https://github.com/xrcodexcode/xr_node --skill youtube-ingestion-agent-xrcodexcode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: youtube-ingestion-agent
Source: https://github.com/xrcodexcode/xr_node/tree/main/claude/skills/youtube-ingestion-agent
Command: npx skills add https://github.com/xrcodexcode/xr_node --skill youtube-ingestion-agent-xrcodexcode

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Raw YouTube transcripts are messy, full of ASR artifacts and code-switched language, and turning them into clean, schema-compliant study notes for a knowledge base is tedious manual work. ## Core Features & Use Cases - Transcript Cleanup & Translation: Removes ASR artifacts and translates non-English or code-switched phrases into professional English. - Schema-Compliant Note Generation: Produces exactly one study note validated against Frontmatter Schema v4, UUID v4, and a controlled tag vocabulary. - Approval-Gated Lifecycle: Stages drafts in a processing folder, pauses for explicit user approval, then promotes the note and archives the raw source. - Use Case: After saving a transcript of a recorded lecture into the capture folder, invoke this Skill to receive a polished study note with Mermaid diagrams, ready for promotion into your knowledge base. ## Quick Start Ask the agent to ingest the raw YouTube transcript waiting in the capture folder and turn it into a study note for review.

Frequently Asked Questions about youtube-ingestion-agent

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

FAQPage Schema
How do I turn a YouTube transcript into a study note?

Place the raw transcript in the capture folder and invoke the ingestion agent. It cleans ASR artifacts, translates code-switched phrases, and produces one schema-validated study note draft for your approval.

How to handle code-switched or non-English YouTube transcripts?

The ingestion workflow translates non-English and code-switched phrases into professional English during the processing stage. The resulting note uses consistent English while preserving the original meaning of the transcript.

Does the ingestion process modify the original transcript file?

No, the capture folder is strictly read-only until explicit user approval. A working draft is created in the processing folder, and the raw transcript is only archived after you approve the final note.

What metadata validation happens before a note is published?

The draft is validated against Frontmatter Schema v4, UUID v4 identifiers, and a controlled tag vocabulary defined in the tagging rules. The agent stops and waits for user approval before promoting the note.

Can the agent generate multiple notes from one transcript?

No, the workflow produces exactly one comprehensive study note per transcript. It also enforces a zero-hallucination policy, never inventing facts or citations absent from the source transcript.