MCP Prompts Reference
Prompts are pre-built research workflows that combine multiple data sources into a structured analysis request. Select a prompt in your AI client to start a guided research session.
research_video
Comprehensive analysis of a single video.
| Argument | Type | Description |
|---|---|---|
videoId | string | YouTube video ID to research |
The prompt automatically fetches and combines:
- Video metadata (title, description, dates, status)
- Transcript excerpt (first 2000 characters)
- AI summary (if available)
- Status change history
Your AI assistant then analyzes this combined context to provide a comprehensive report.
Example Use Cases
- Investigating why a video was deleted
- Summarizing a long lecture or documentary
- Fact-checking claims in a video
channel_overview
Research a YouTube channel's history and content.
| Argument | Type | Description |
|---|---|---|
channelId | string | YouTube channel ID |
The prompt fetches:
- Channel metadata (name, description, subscriber count)
- Video count by status (live vs. deleted)
- Most recent 10 video titles
- Summary statistics
Example Use Cases
- Researching a content creator's history
- Analyzing a terminated channel's archive
- Understanding content patterns
find_deleted
Discover and analyze deleted videos on a topic.
| Argument | Type | Description |
|---|---|---|
query | string | Search topic or keywords |
The prompt runs a full-text search filtered to deleted videos and returns the top 10 results with guidance on deeper analysis using get_subtitles and get_summary.
Example Use Cases
- Finding deleted music videos or performances
- Researching censored or removed content
- Discovering historical videos no longer on YouTube