MCP YouTube Intelligence

MCP YouTube Intelligence

An MCP server for intelligent YouTube video analysis that provides token-optimized summaries, sentiment analysis, and entity extraction from transcripts. It enables AI assistants to perform video reporting, channel monitoring, and comprehensive YouTube searches through structured data tools.

Category
Visit Server

README

Python License MCP PyPI

🌐 English | ν•œκ΅­μ–΄

MCP YouTube Intelligence

YouTube μ˜μƒμ„ μ§€λŠ₯적으둜 λΆ„μ„ν•˜λŠ” MCP μ„œλ²„ + CLI

MCP (Model Context Protocol)λŠ” Claude, Cursor 같은 AI 도ꡬ가 μ™ΈλΆ€ μ„œλΉ„μŠ€λ₯Ό μ‚¬μš©ν•  수 있게 ν•΄μ£ΌλŠ” ν‘œμ€€ ν”„λ‘œν† μ½œμž…λ‹ˆλ‹€. 이 μ„œλ²„λ₯Ό μ—°κ²°ν•˜λ©΄ "이 μ˜μƒ μš”μ•½ν•΄μ€˜" ν•œλ§ˆλ””λ‘œ 뢄석이 μ™„λ£Œλ©λ‹ˆλ‹€.

🎯 핡심 κ°€μΉ˜: 원본 μžλ§‰(2,000~30,000 토큰)을 μ„œλ²„μ—μ„œ μ²˜λ¦¬ν•˜μ—¬ LLMμ—λŠ” ~200–500 ν† ν°λ§Œ μ „λ‹¬ν•©λ‹ˆλ‹€.


πŸ€” μ™œ 이 μ„œλ²„μΈκ°€?

λŒ€λΆ€λΆ„μ˜ YouTube MCP μ„œλ²„λŠ” 원본 μžλ§‰μ„ κ·ΈλŒ€λ‘œ LLM에 λ˜μ§‘λ‹ˆλ‹€.

κΈ°λŠ₯ κΈ°μ‘΄ MCP μ„œλ²„ MCP YouTube Intelligence
μžλ§‰ μΆ”μΆœ βœ… βœ…
μ„œλ²„μ‚¬μ΄λ“œ μš”μ•½ (토큰 μ΅œμ ν™”) ❌ βœ…
κ΅¬μ‘°ν™”λœ 리포트 (μš”μ•½+ν† ν”½+μ—”ν‹°ν‹°+λŒ“κΈ€) ❌ βœ…
채널 λͺ¨λ‹ˆν„°λ§ (RSS) ❌ βœ…
λŒ“κΈ€ 감성 뢄석 ❌ βœ…
ν† ν”½ μ„Έκ·Έλ©˜ν…Œμ΄μ…˜ ❌ βœ…
μ—”ν‹°ν‹° μΆ”μΆœ (ν•œ/영 200+개) ❌ βœ…
μžλ§‰/YouTube 검색 ❌ βœ…
배치 처리 ❌ βœ…
SQLite/PostgreSQL μΊμ‹œ ❌ βœ…

πŸš€ λΉ λ₯Έ μ‹œμž‘

1. μ„€μΉ˜

pip install mcp-youtube-intelligence
pip install yt-dlp  # μžλ§‰ μΆ”μΆœμ— ν•„μš”

πŸ’‘ LLM 없이도 κΈ°λ³Έ μš”μ•½(핡심 λ¬Έμž₯ μΆ”μΆœ)은 λ™μž‘ν•©λ‹ˆλ‹€. κ³ ν’ˆμ§ˆ μš”μ•½μ„ μ›ν•˜λ©΄ μ•„λž˜ LLM 섀정을 μ°Έκ³ ν•˜μ„Έμš”.

2. 첫 번째 λͺ…λ Ήμ–΄ μ‹€ν–‰

# 리포트 생성 β€” μš”μ•½, ν† ν”½, μ—”ν‹°ν‹°, λŒ“κΈ€μ„ ν•œλ²ˆμ— 뢄석 (LLM μ—°λ™ν•„μš”)
mcp-yt report "https://www.youtube.com/watch?v=LV6Juz0xcrY"

# μžλ§‰ μš”μ•½λ§Œ
mcp-yt transcript "https://www.youtube.com/watch?v=LV6Juz0xcrY"

# μ˜μƒ ID만 써도 λ©λ‹ˆλ‹€
mcp-yt report LV6Juz0xcrY

⚠️ zsh μ‚¬μš©μž: URL에 ?κ°€ μžˆμœΌλ―€λ‘œ λ°˜λ“œμ‹œ λ”°μ˜΄ν‘œλ‘œ κ°μ‹Έμ„Έμš”.

πŸ“‹ 리포트 좜λ ₯ μ˜ˆμ‹œ

mcp-yt report "https://www.youtube.com/watch?v=LV6Juz0xcrY" μ‹€ν–‰ κ²°κ³Ό (extractive μš”μ•½):

# πŸ“Ή Video Analysis Report: OpenClaw Use Cases that are Actually Helpful! (ClawdBot)

> Channel: Duncan Rogoff | AI Automation | Duration: 16:29 | Language: en_ytdlp

## 1. Summary

OpenClaw is the most powerful AI agent framework in the world right now and
it's about to replace your entire workflow. I spent over $200 in the last
48 hours stress testing the system so you don't have to. It defines who it
is, how it behaves, and crucial behavioral boundaries. If you think open
claw is cool, just check out this video up here of 63 insane use cases
that other people are doing.

## 2. Key Topics

| # | Topic | Keywords | Timespan |
|---|-------|----------|----------|
| 1 | framework, world, right | framework, world, right | 0:00~0:05 |
| 2 | like, really, there | like, really, there | 0:05~2:23 |
| 3 | like, max, using | like, max, using | 2:23~4:22 |
| 4 | going, like, something | going, like, something | 4:22~5:03 |
| 5 | like, agents, basically | like, agents, basically | 5:03~6:04 |
| ... | ... | ... | ... |
| 15 | think, open, claw | think, open, claw | 16:24~16:29 |

## 4. Keywords & Entities

- **Technology**: GitHub, LLM, GPT
- **Company**: Anthropic, Apple

## 5. Viewer Reactions

- Total comments: 20
- Sentiment: Positive 45% / Negative 0% / Neutral 55%
- Top opinions:
  - **@geetee2583** (positive, πŸ‘8): Great info. Just need your inset video out of the way...
  - **@bdog4026** (positive, πŸ‘3): This tool is wild! Definitely the most in depth explanation...
  - **@magalyvilela4917** (neutral, πŸ‘3): Came to this video wondering it gonna teach me how to set up...

πŸ“– CLI 전체 λͺ…λ Ήμ–΄

πŸ“Š 리포트 (핡심 κΈ°λŠ₯)

⚠️ **리포트의 μš”μ•½ μ„Ήμ…˜μ€ LLM 연동이 ν•„μˆ˜μž…λ‹ˆλ‹€. Ollama λΉ λ₯Έ μ„€μ • (무료, 3뢄이면 끝):

# 1. Ollama μ„€μΉ˜: https://ollama.ai
# 2. λͺ¨λΈ λ‹€μš΄λ‘œλ“œ
ollama pull qwen2.5:7b

# 3. ν™˜κ²½λ³€μˆ˜ μ„€μ •
export MYI_LLM_PROVIDER=ollama
export MYI_OLLAMA_MODEL=qwen2.5:7b

# 원격 μ„œλ²„λΌλ©΄ ν˜ΈμŠ€νŠΈλ„ μ§€μ •
export MYI_OLLAMA_BASE_URL=http://your-server:11434
mcp-yt report "https://youtube.com/watch?v=VIDEO_ID"
mcp-yt report VIDEO_ID --provider ollama     # LLM ν”„λ‘œλ°”μ΄λ” μ§€μ •
mcp-yt report VIDEO_ID --no-comments         # λŒ“κΈ€ μ œμ™Έ
mcp-yt report VIDEO_ID -o report.md          # 파일 μ €μž₯

🎯 μžλ§‰ μΆ”μΆœ + μš”μ•½

mcp-yt transcript VIDEO_ID                   # μš”μ•½ (~200–500 토큰)
mcp-yt transcript VIDEO_ID --mode full       # 전체 μžλ§‰
mcp-yt transcript VIDEO_ID --mode chunks     # 청크 λΆ„ν• 
mcp-yt --json transcript VIDEO_ID            # JSON 좜λ ₯

기타

mcp-yt video VIDEO_ID                        # 메타데이터
mcp-yt comments VIDEO_ID --max 20            # λŒ“κΈ€ (감성 뢄석 포함)
mcp-yt entities VIDEO_ID                     # μ—”ν‹°ν‹° μΆ”μΆœ
mcp-yt segments VIDEO_ID                     # ν† ν”½ μ„Έκ·Έλ©˜ν…Œμ΄μ…˜
mcp-yt search "ν‚€μ›Œλ“œ" --max 5               # YouTube 검색
mcp-yt monitor subscribe @채널핸듀           # 채널 λͺ¨λ‹ˆν„°λ§
mcp-yt playlist PLAYLIST_ID                  # ν”Œλ ˆμ΄λ¦¬μŠ€νŠΈ
mcp-yt batch ID1 ID2 ID3                     # 배치 처리
mcp-yt search-transcripts "ν‚€μ›Œλ“œ"           # μ €μž₯된 μžλ§‰ 검색

πŸ’‘ λͺ¨λ“  λͺ…령어에 --json ν”Œλž˜κ·Έλ₯Ό μΆ”κ°€ν•˜λ©΄ JSON 좜λ ₯λ©λ‹ˆλ‹€.


πŸ”Œ MCP μ„œλ²„ μ—°κ²°

MCP μ„œλ²„λŠ” stdio ν”„λ‘œν† μ½œλ‘œ ν†΅μ‹ ν•©λ‹ˆλ‹€.

Claude Desktop / Cursor / OpenCode

μ„€μ • νŒŒμΌμ— μΆ”κ°€ (claude_desktop_config.json, .cursor/mcp.json, mcp.json):

{
  "mcpServers": {
    "youtube": {
      "command": "uvx",
      "args": ["mcp-youtube-intelligence"],
      "env": {
        "MYI_LLM_PROVIDER": "ollama",
        "MYI_OLLAMA_MODEL": "qwen2.5:7b"
      }
    }
  }
}

πŸ’‘ uvxλŠ” uv νŒ¨ν‚€μ§€ λ§€λ‹ˆμ €μ˜ μ‹€ν–‰ λͺ…λ Ήμ–΄μž…λ‹ˆλ‹€. pip install uv둜 μ„€μΉ˜ν•˜μ„Έμš”.

ν΄λΌμš°λ“œ LLM을 μ“°λ €λ©΄ env에 API ν‚€λ₯Ό μΆ”κ°€ν•˜λ©΄ λ©λ‹ˆλ‹€: "OPENAI_API_KEY": "sk-..."

Claude Code

claude mcp add youtube -- uvx mcp-youtube-intelligence

MCP Tools (9개)

Tool μ„€λͺ… μ˜ˆμƒ 토큰
get_video 메타데이터 + μš”μ•½ ~200–500
get_transcript μžλ§‰ (summary/full/chunks) ~200–500
get_comments λŒ“κΈ€ + 감성 뢄석 ~200–500
monitor_channel RSS 채널 λͺ¨λ‹ˆν„°λ§ ~100–300
search_transcripts μ €μž₯된 μžλ§‰ 검색 ~100–400
extract_entities μ—”ν‹°ν‹° μΆ”μΆœ ~150–300
segment_topics ν† ν”½ λΆ„ν•  ~100–250
search_youtube YouTube 검색 ~200
get_playlist ν”Œλ ˆμ΄λ¦¬μŠ€νŠΈ 뢄석 ~200–500

<details> <summary>πŸ“– Tool νŒŒλΌλ―Έν„° 상세</summary>

get_video

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ μ„€λͺ…
video_id string βœ… YouTube μ˜μƒ ID

get_transcript

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
video_id string βœ… β€” YouTube μ˜μƒ ID
mode string ❌ "summary" summary · full · chunks

get_comments

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
video_id string βœ… β€” YouTube μ˜μƒ ID
top_n int ❌ 10 λ°˜ν™˜ν•  λŒ“κΈ€ 수
summarize bool ❌ false μš”μ•½ λ·°

monitor_channel

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
channel_ref string βœ… β€” 채널 URL/@ν•Έλ“€/ID
action string ❌ "check" add·check·list·remove

search_transcripts

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
query string βœ… β€” 검색 ν‚€μ›Œλ“œ
limit int ❌ 10 μ΅œλŒ€ κ²°κ³Ό 수

extract_entities / segment_topics

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ μ„€λͺ…
video_id string βœ… YouTube μ˜μƒ ID

search_youtube

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
query string βœ… β€” 검색 ν‚€μ›Œλ“œ
max_results int ❌ 10 μ΅œλŒ€ κ²°κ³Ό 수
order string ❌ "relevance" relevance·date·rating·viewCount

get_playlist

νŒŒλΌλ―Έν„° νƒ€μž… ν•„μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
playlist_id string βœ… β€” ν”Œλ ˆμ΄λ¦¬μŠ€νŠΈ ID
max_videos int ❌ 50 μ΅œλŒ€ μ˜μƒ 수

</details>


βš™οΈ μ„€μ •

LLM ν”„λ‘œλ°”μ΄λ” μ„€μ •

LLM 없이도 κΈ°λ³Έ μš”μ•½(핡심 λ¬Έμž₯ μΆ”μΆœ)은 λ™μž‘ν•©λ‹ˆλ‹€. κ³ ν’ˆμ§ˆ μš”μ•½μ„ μ›ν•˜λ©΄:

Ollama (μΆ”μ²œ β€” 무료, μ˜€ν”„λΌμΈ)

# 1. Ollama μ„€μΉ˜: https://ollama.ai
# 2. λͺ¨λΈ λ‹€μš΄λ‘œλ“œ
ollama pull qwen2.5:7b

# 3. ν™˜κ²½λ³€μˆ˜ μ„€μ •
export MYI_LLM_PROVIDER=ollama
export MYI_OLLAMA_MODEL=qwen2.5:7b

# 4. (선택) 원격 Ollama μ„œλ²„ μ‚¬μš© μ‹œ
export MYI_OLLAMA_BASE_URL=http://your-server:11434

ν΄λΌμš°λ“œ LLM

# API ν‚€λ§Œ μ„€μ •ν•˜λ©΄ μžλ™ 감지 (MYI_LLM_PROVIDER=auto)
export OPENAI_API_KEY=sk-...          # OpenAI
export ANTHROPIC_API_KEY=sk-ant-...   # Anthropic
export GOOGLE_API_KEY=AIza...         # Google

# νŠΉμ • ν”„λ‘œλ°”μ΄λ” μ§€μ •
export MYI_LLM_PROVIDER=anthropic

ν΄λΌμš°λ“œ LLM νŒ¨ν‚€μ§€: pip install "mcp-youtube-intelligence[llm]" (OpenAI) / [anthropic-llm] / [google-llm] / [all-llm]

μΆ”μ²œ Ollama λͺ¨λΈ

λͺ©μ  λͺ¨λΈ 크기 ν•œκ΅­μ–΄ μ˜μ–΄ ν’ˆμ§ˆ
λ‹€κ΅­μ–΄ (μΆ”μ²œ) qwen2.5:7b 4.4GB βœ… βœ… ⭐⭐⭐⭐
μ˜μ–΄ 쀑심 llama3.1:8b 4.7GB ⚠️ βœ… ⭐⭐⭐⭐
ν•œκ΅­μ–΄ νŠΉν™” gemma2:9b 5.4GB βœ… βœ… ⭐⭐⭐⭐
κ²½λŸ‰ qwen2.5:3b 1.9GB βœ… βœ… ⭐⭐⭐
λ‹€κ΅­μ–΄ νŠΉν™” aya-expanse:8b 4.8GB βœ… βœ… ⭐⭐⭐

⏱️ μ‹€μΈ‘ 벀치마크

RTX 3070 8GB Β· Ollama Β· ν•œκ΅­μ–΄ μžλ§‰ ~2,900자 (5λΆ„ 19초 μ˜μƒ) load_duration μ œμ™Έ, 순수 생성 μ‹œκ°„ κΈ°μ€€

λͺ¨λΈ Prompt 처리 생성 μ‹œκ°„ 속도 좜λ ₯ ν’ˆμ§ˆ
Extractive - μ¦‰μ‹œ - 379자 ⭐⭐
qwen2.5:1.5b 7.8s 4.7s 30.4 tok/s 232자 ⭐⭐
qwen2.5:7b 34.5s 18.8s 7.3 tok/s 766자 ⭐⭐⭐⭐
aya-expanse:8b 29.5s 34.5s 6.2 tok/s 405자 ⭐⭐⭐

⚠️ 첫 μ‹€ν–‰ μ‹œ λͺ¨λΈ λ‘œλ“œμ— 15~60초 μΆ”κ°€. keep_alive둜 λ©”λͺ¨λ¦¬ μœ μ§€ν•˜λ©΄ 이후 λ‘œλ“œ μ—†μŒ.

<details> <summary>πŸ“– 전체 ν™˜κ²½λ³€μˆ˜ λͺ©λ‘</summary>

ν™˜κ²½λ³€μˆ˜ κΈ°λ³Έκ°’ μ„€λͺ…
MYI_DATA_DIR ~/.mcp-youtube-intelligence 데이터 디렉토리
MYI_STORAGE sqlite sqlite Β· postgres
MYI_SQLITE_PATH {DATA_DIR}/data.db SQLite 경둜
MYI_POSTGRES_DSN β€” PostgreSQL DSN
MYI_YT_DLP yt-dlp yt-dlp 경둜
MYI_MAX_COMMENTS 20 μ΅œλŒ€ λŒ“κΈ€ 수
MYI_LLM_PROVIDER auto autoΒ·openaiΒ·anthropicΒ·googleΒ·ollamaΒ·vllmΒ·lmstudio
OPENAI_API_KEY β€” OpenAI ν‚€
MYI_OPENAI_MODEL gpt-4o-mini OpenAI λͺ¨λΈ
ANTHROPIC_API_KEY β€” Anthropic ν‚€
MYI_ANTHROPIC_MODEL claude-sonnet-4-20250514 Anthropic λͺ¨λΈ
GOOGLE_API_KEY β€” Google ν‚€
MYI_GOOGLE_MODEL gemini-2.0-flash Google λͺ¨λΈ
MYI_OLLAMA_BASE_URL http://localhost:11434 Ollama URL
MYI_OLLAMA_MODEL llama3.1:8b Ollama λͺ¨λΈ
MYI_VLLM_BASE_URL http://localhost:8000 vLLM URL
MYI_VLLM_MODEL β€” vLLM λͺ¨λΈ
MYI_LMSTUDIO_BASE_URL http://localhost:1234 LM Studio URL
MYI_LMSTUDIO_MODEL β€” LM Studio λͺ¨λΈ

</details>


πŸ” νŠΈλŸ¬λΈ”μŠˆνŒ…

문제 ν•΄κ²°
zsh: no matches found URL을 λ”°μ˜΄ν‘œλ‘œ 감싸기: mcp-yt transcript "https://..."
yt-dlp not found pip install yt-dlp λ˜λŠ” export MYI_YT_DLP=/path/to/yt-dlp
μžλ§‰ μ—†λŠ” μ˜μƒ mcp-yt video둜 λ©”νƒ€λ°μ΄ν„°λ§Œ κ°€μ Έμ˜€κΈ°
SQLite database locked μ„œλ²„ μΈμŠ€ν„΄μŠ€ ν•˜λ‚˜λ§Œ μ‹€ν–‰ 쀑인지 확인
LLM μš”μ•½ μ‹€νŒ¨ μžλ™μœΌλ‘œ extractive 폴백됨. API ν‚€ 확인.

🀝 Contributing

git clone https://github.com/JangHyuckYun/mcp-youtube-intelligence.git
cd mcp-youtube-intelligence
pip install -e ".[dev]"
pytest tests/ -v

πŸ“„ λΌμ΄μ„ μŠ€

Apache 2.0 β€” LICENSE

πŸ“ λ³€κ²½ 이λ ₯

λ‚ μ§œ 버전 μ£Όμš” λ³€κ²½
2025-02-18 v0.1.0 초기 릴리슀 β€” 9개 MCP 도ꡬ, CLI, SQLite
2025-02-18 v0.1.1 Multi-LLM (OpenAI/Anthropic/Google), Apache 2.0
2025-02-18 v0.1.2 Local LLM (Ollama/vLLM/LM Studio), yt-dlp μžλ§‰ κ°œμ„ , μ˜μ–΄ κΈ°λ³Έ 좜λ ₯
2025-02-18 v0.1.3 Local LLM (Ollama/vLLM/LM Studio), yt-dlp μžλ§‰ κ°œμ„ , μ˜μ–΄ κΈ°λ³Έ 좜λ ₯

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured