# kirbah-mcp-youtube [Health: Active]

**Category:** 🗄️ Databases  
**Repository:** https://github.com/kirbah/mcp-youtube  
**GitHub Stars:** 29  
**npm Downloads (last month):** 2119  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/kirbah-mcp-youtube

## Description
Provide token-optimized, structured YouTube data to enhance your LLM applications. Access efficien…

## Tools
Capabilities this server exposes over MCP:

- **getVideoDetails** — Retrieves detailed, **lean** information for multiple YouTube videos including metadata, statistics, engagement ratios, and content details.
- **searchVideos** — Searches for videos or channels based on a query string with various filtering options, returning **concise** results.
- **getTranscripts** — Retrieves **token-efficient** transcripts (captions) for multiple videos, with options for full text or key segments (intro/outro).
- **getChannelStatistics** — Retrieves **lean** statistics for multiple channels (subscriber count, view count, video count, creation date).
- **getChannelTopVideos** — Retrieves a list of a channel's top-performing videos with **lean** details and engagement ratios.
- **getTrendingVideos** — Retrieves a list of trending videos for a given region and optional category, with **lean** details and engagement ratios.
- **getVideoCategories** — Retrieves available YouTube video categories (ID and title) for a specific region, providing **essential data only**.
- **getVideoComments** — Retrieves comments for a YouTube video. Allows sorting, limiting results, and fetching a small number of replies per comment.
- **findConsistentOutlierChannels** — Identifies channels that consistently perform as outliers within a specific niche. **Requires a MongoDB connection.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "kirbah-mcp-youtube": {
    "command": "npx",
    "args": ["-y","@kirbah/mcp-youtube"]
  }
}
```

## Documentation

## What kirbah-mcp-youtube does

kirbah-mcp-youtube MCP server gives an MCP client access to selected YouTube data without returning the full, verbose shape of typical API responses. Its tools cover video metadata and statistics, search, captions, channel metrics, top-performing videos, trends, categories, and comments. Results are structured for language-model processing, with an emphasis on concise fields such as views, likes, engagement ratios, titles, dates, and content details.

The server also includes a niche-analysis operation that identifies channels that consistently perform as outliers. That operation requires a MongoDB connection. The README describes MongoDB as an optional caching layer as well, allowing repeated requests to be served from stored results rather than consuming additional YouTube API quota.

## How it works

The package runs as a local stdio MCP server through the npm package `@kirbah/mcp-youtube`. An MCP-compatible client starts it with `npx`, then exposes the registered tools to an agent. Requests can contain multiple video or channel IDs where supported, which is useful for comparing items in one call.

Transcript retrieval is available in zero-configuration mode. Captions can be requested as full text or as selected segments such as introductions and conclusions, and the README notes support for multiple languages. Other capabilities depend on the YouTube Data API v3 key. MongoDB is optional but supports caching and is needed for consistent outlier-channel analysis.

## Setup and configuration

Install the kirbah-mcp-youtube MCP server with the following client configuration:

```json
{
  "mcpServers": {
    "youtube": {
      "command": "npx",
      "args": ["-y", "@kirbah/mcp-youtube"]
    }
  }
}
```

Add `YOUTUBE_API_KEY` when the client needs the YouTube Data API-backed search and analytics features. Add `MDB_MCP_CONNECTION_STRING` to enable MongoDB-backed caching and the outlier-channel analysis. The README presents a MongoDB connection as recommended for quota savings, not as a prerequisite for basic transcript use.

The documented manual setup applies to clients such as Claude Desktop and Cursor. The project also provides an “Add to Claude Desktop” option through its Glama listing, but the local npm configuration is the direct installation method described in the README.

## Tools and capabilities

The kirbah-mcp-youtube MCP server exposes these operations:

- `getVideoDetails` returns lean metadata, statistics, engagement ratios, and content details for multiple videos.
- `searchVideos` searches videos or channels with query, result-count, ordering, type, channel, and related filters.
- `getTranscripts` retrieves captions for multiple videos as full text or selected segments.
- `getChannelStatistics` returns subscriber, view, video-count, and creation-date information.
- `getChannelTopVideos` lists a channel’s strongest-performing videos with concise details and engagement ratios.
- `getTrendingVideos` finds trending videos by region and optional category.
- `getVideoCategories` lists category IDs and titles for a region.
- `getVideoComments` retrieves comments with sorting, result limits, and a small number of replies per comment.
- `findConsistentOutlierChannels` identifies consistently unusual channel performance within a niche and requires MongoDB.

## Limitations and notes

The server depends on YouTube data availability and, for API-backed operations, the YouTube Data API v3 quota and credentials. Transcript availability depends on whether captions exist for the requested video. MongoDB caching is optional, so deployments without it do not receive that cache-based quota benefit. The README reports token-reduction measurements for selected operations, but those figures should be treated as the project’s examples rather than a guarantee for every request.

## Getting started with this kirbah-mcp-youtube MCP server
Always refer to the official documentation for the most accurate and up-to-date information.

_Full upstream README: https://allmcps.com/mcp/kirbah-mcp-youtube/readme_