youtube-collector

You can automatically collect new YouTube videos and generate transcript summaries by registering channel handles and triggering an end-to-end workflow that fetches latest videos, pulls transcripts, and creates summaries.

Yes, fetching video transcripts via the YouTube API and storing them in YAML format is fully supported. Summaries and video metadata are saved as YAML files under the .reference directory for structured access.

Yes, a YouTube API key is required to register channels and collect transcripts. The workflow enforces API key configuration and handles common errors like missing keys, unregistered channels, or API quota issues.

Managing multiple YouTube channels and avoiding duplicate video collection is handled natively by the workflow. It automatically skips already collected videos, ensuring only new video transcripts are processed and summarized.

YouTube transcript collection fails when the configured API key is missing, the channel is unregistered, or API quota limits are exceeded. The workflow explicitly checks for these common API errors during channel lookup and video retrieval.

Transcript-based summarization works by extracting text from fetched video transcripts and generating concise summaries. These summaries are then stored alongside video data in YAML files within the .reference directory.