YouTube Media Integration Module

The YouTube MCP module facilitates the retrieval of transcribed content directly from video assets hosted on YouTube.

Operational Interfaces

retrieve_video_captions

Fetches the associated caption text for a given YouTube video identifier.

Arguments Required: - sourceUrl: The complete Uniform Resource Locator for the YouTube content (accepts standard links, short domains, and embedded player paths).

Output Structure: - A JSON Object containing: - data_array: A collection holding the segmented transcript lines.

Supported Link Schemas

Processing logic recognizes several standard YouTube addressing schemes: - Canonical: https://www.youtube.com/watch?v=IDENTIFIER - Shortened: https://youtu.be/IDENTIFIER - Embedding: https://www.youtube.com/embed/IDENTIFIER

Exception Management

Robust error handling is implemented: - Malformed or nonexistent URLs trigger ErrorCode.InvalidParams - Omission of the URL argument results in ErrorCode.InvalidParams - Failures during external data retrieval yield descriptive error messages - Ensures clean termination upon receipt of SIGINT signal

Implementation Specifics

• Developed leveraging the Highlight AI MCP Software Development Kit
• Dependent upon the 'youtube-transcript' underlying library
• Input schema validation is enforced using Zod
• Operates as a standard input/output communication server process
• Requires a runtime environment of Node.js version 18.0.0 or newer

Constraints

• Functionality is contingent upon captions being actively enabled for the specific video.
• Currently restricted to extracting transcripts in the English language only.
• Throttling limits are governed by YouTube's public access policies.