# youtube-knowledge-extractor

> youtube-knowledge-extractor — sdrabent-youtube-knowledge-extractor. Use this tool when you need to analyze YouTube videos and extract relevant information from both audio and visual content. It solves problems such as sentiment analysis, object detection, and topic modeling by processing video transcripts and frame-level image analysis. The tool takes YouTube video URLs as input and outputs extracted knowledge in the form of text summaries, entity recognitions, and visual insights.

Canonical page: https://skillsregistry.net/skills/sdrabent-youtube-knowledge-extractor  
JSON: https://api.skillsregistry.net/v1/skills/sdrabent-youtube-knowledge-extractor

## Description

Multimodal YouTube video analysis through both audio (transcript) and visual (frame extraction + image analysis)

## Trust

- **Trust score (0–1):** 0.65
- **Verification tier:** scanned
- **Last scanned:** 2026-05-19

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** instructions
- **Runtime environment:** llm
- **Category:** social-media
- **Updated:** 2026-09-20

## Source

- **Source listing:** [ClawHub](https://clawskills.sh/skills/sdrabent-youtube-knowledge-extractor)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "sdrabent-youtube-knowledge-extractor"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/sdrabent-youtube-knowledge-extractor` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/sdrabent-youtube-knowledge-extractor/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt
