Turns one YouTube podcast episode into a fact-checked, bilingual (RU/EN), interactive self-contained HTML deep-dive — fetches the transcript via yt-dlp, runs a multi-agent analysis (map → per-section analyze → adversarial verify → synthesize) into a single podcast.json source of truth with speakers, topic sections, concepts, attributed speaker quotes, sparse debate cards and deep-link timestamps, grounds every quote against the transcript, translates RU↔EN, and renders a page with inline-SVG/CSS visuals, a chapter timeline, a searchable glossary and active-recall Q&A. Use when the user wants a podcast deep-dive, to analyze/break down a YouTube episode, or to package an episode into an interactive lesson page.
$npx -y skills add mldogs/skill-factory --skill podcast-deepdiveInstalls into the current project.
Run `npx skills use "https://github.com/mldogs/skill-factory" --skill "mldogs/skill-factory/podcast-deepdive"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.
Use the skills in "https://github.com/mldogs/skill-factory" that are relevant to the current task. Run `npx skills add "https://github.com/mldogs/skill-factory"` and select the relevant skills, then follow their instructions.