youtube-transcript-browser

name	youtube-transcript-browser
description	Use browser-backed YouTube adapters to pull transcripts, video metadata, and discussion context from market interviews, conference clips, podcasts, and earnings commentary.
metadata	{"marketbot":{"emoji":"▶️","triggers":["youtube transcript","video transcript","earnings interview","podcast transcript","youtube video"],"output":"youtube-transcript-report","risk":"low","freshness":"live","tools":["browser_site"],"required_tools":["browser_site"],"markets":["global","mixed"],"asset_classes":["equity","crypto","commodity","macro","etf"],"task_type":"browser-research","determinism":"tool-backed","priority":84}}

YouTube Transcript Browser

Use this skill when the user needs transcript-first analysis from a market video, podcast, interview, conference clip, or earnings discussion.

Workflow

1. Use browser_site with YouTube adapters that exist in the runtime catalog. Prefer exact adapters such as:
   • youtube/transcript
   • youtube/search
   • youtube/video
2. Read references/adapter-examples.md when you need concrete adapter call patterns or fallback behavior.
3. Extract:
   • key claims
   • management or speaker tone
   • forward-looking guidance or thesis points
4. Pair with summarize or earnings-readout when the transcript needs a more structured write-up.

Rules

• Prefer youtube/transcript when available in the catalog; only fall back to youtube/video or youtube/search when transcript access is unavailable.
• Do not invent adapter names like youtube/comments unless the runtime catalog explicitly exposes them.
• Prefer transcript over title-only interpretation.
• Flag when transcript quality looks incomplete or auto-generated.