YouTube Video Transcript

Retrieve a YouTube video transcript through Frevana.

Purpose

This skill calls Frevana's /service/youtube-video-transcript endpoint and returns the SerpAPI response unchanged.

Required input: v.

Supported request fields:

• v (required)
• language_code
• title
• type
• no_cache
• async
• zero_trace

The script validates that the response is JSON and prints it to stdout. Do not rewrite or reshape returned data unless the user explicitly asks for a transformation.

What This Skill Needs

• YouTube video ID
• FREVANA_TOKEN in the environment variables, or --token for the current run
• curl, bash, and python3

The API base URL defaults to https://ai-factory.frevana.com. Set FREVANAAPIBASE_URL to point at another host.

Commands

Get a transcript

bash <skill-path>/scripts/get_youtube_video_transcript.sh \
  --v "Gk8gB5VACZw" \
  --language-code "en"

Save JSON to a file

bash <skill-path>/scripts/get_youtube_video_transcript.sh \
  --v "V" \
  --output "./out/youtube-video-transcript.json"

Fixed Request Shape

The script sends this payload shape, omitting optional fields that were not provided:

{
  "v": "V",
  "language_code": "LANGUAGE_CODE",
  "title": "TITLE",
  "type": "TYPE",
  "no_cache": true,
  "async": true,
  "zero_trace": true
}

Do not pass unsupported fields. Do not pass SerpAPI engine or api_key; Frevana sets those server-side. Use --output for local file saving.

Output

• Success: validated JSON on stdout
• With --output: the same JSON is also written to the requested file
• Failure: response body or parsing error on stderr, with a non-zero exit code

Example Prompts

• "Get YouTube video transcript for YouTube video ID."
• "Run YouTube Video Transcript and save the raw JSON."