YouTube Video Finder Instructions

Use this skill whenever a user asks to find, recover, or check the archive status of a deleted, missing, or private YouTube video.

Execution Steps:

1. Extract the Video ID: Isolate the 11-character ID from the URL or text provided by the user (e.g., for youtube.com/watch?v=dQw4w9WgXcQ, the ID is dQw4w9WgXcQ). Pass only this ID into the videoid parameter.
2. Parameters: Omit the includeRaw and stream parameters unless the user specifically asks for raw metadata or debugging data.

Interpreting the Response:

• Initial Check: Look at the status field. If it returns bad.id, stop and inform the user that the 11-character ID is invalid.
• The Verdict: Check the verdict object. Read the human_friendly string to immediately understand the overall result. Pay attention to the boolean flags video, metaonly, and comments to know exactly what was recovered.
• Finding the Links: Iterate through the keys array (which contains service objects like GhostArchive or Wayback Machine).
  • If a service object has archived: true, look inside its available list.
  • Extract the url from the Link Objects to give to the user.
  • Take note of maybe_paywalled (boolean) or note (string) for context on how to access the link.

Output Formatting:

Present your findings in a clear, friendly, and structured manner.

• If the video is fully found, provide the direct links immediately.
• If only metadata or comments were archived, clarify that the actual video file could not be recovered.
• Mention which service(s) successfully held the archive.