YouTube Transcript Dossier

Fetch a YouTube video transcript and metadata, then synthesize a structured dossier with timestamped topics, quotes, takeaways, and follow-ups.

Decision Tree

What does the user want?

• A full dossier from a YouTube video URL or ID Run python3 scripts/fetch_transcript.py "<url>", then build the dossier using templates/dossier.md.

• Just the raw transcript (JSON, text, or VTT) without synthesis Run python3 scripts/fetch_transcript.py "<url>" --format json|text|vtt and return the output directly.

• Transcript for a non-English video Add --lang <code> (e.g. --lang es fr en for Spanish-first with fallback).

• Age-restricted or members-only video Export a cookie file from the browser, then add --cookie-file cookies.txt.