Skip to content

[Bug]: YouTube transcript sometimes extracts only metadata on first attempt #79

Description

@miguelcorderocollar

ExtractMD Version

Unknown

Browser

Unknown

Browser Version

Unknown

Operating System

Unknown

Which integration is affected?

YouTube

Bug Description

When extracting a YouTube transcript, the first attempt sometimes returns only the video metadata, such as title, channel, and URL, with no transcript body. Retrying immediately afterward can return the full transcript.

Steps to Reproduce

  1. Open a YouTube video that has a transcript available.
  2. Run ExtractMD transcript extraction.
  3. Observe that the first extraction may return only metadata with no transcript content.
  4. Run the extraction again on the same page.
  5. Observe that the retry may return the full transcript.

Expected Behavior

The first successful extraction should include the transcript content once the page is ready.

Actual Behavior

The first extraction sometimes captures only metadata and omits the transcript text, even though a retry on the same page can succeed.

Page URL

A YouTube watch page with transcript support.

Console Errors

Not captured yet.

Screenshots

None yet.

Additional Context

This may indicate that YouTube loads transcript content asynchronously in a later phase than basic page metadata, and the extractor can run before the transcript text is fully available. The issue description is intentionally generalized and does not include transcript content from any specific video.

Metadata

Metadata

Assignees

No one assigned

    Labels

    area: extractionContent detection and Markdown extractionbugSomething isn't workingintegration: YouTubeYouTube content extraction

    Projects

    Status
    Backlog

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions