notebooklm-distiller_skill

Overview

This skill automates batch knowledge extraction from Google NotebookLM and writes structured markdown into an Obsidian vault. It supports deep Q&A generation, structured summaries, glossary extraction, web research sessions, and direct markdown persistence. The tool is optimized for repeatable archiving and classroom-style quizzing workflows.

How this skill works

The skill uses NotebookLM CLI sessions to locate or create notebooks, then runs a distillation pipeline that emits QA sets, summaries, or glossaries. It can also start web research sessions to populate a new notebook and persist any generated markdown directly into an Obsidian vault with YAML frontmatter. Optional writeback adds a distill log into the source NotebookLM notebook.

When to use it

• You need a batch export of NotebookLM content into Obsidian (QA, summary, or glossary).
• You want 15–20 deep, source-backed questions for learning or Discord quizzes.
• You want to run web research inside NotebookLM and later extract findings.
• You need to persist meeting notes or generated content into your Obsidian vault.
• You want to push a distillation log back into the original NotebookLM notebook.

Best practices

• Install and authenticate NotebookLM CLI before running commands (notebooklm login).
• Provide explicit --vault-dir and --topic to ensure files land in the right location.
• Choose mode (qa, summary, glossary) based on downstream use: learning, reference, or terminology.
• Use --lang zh only when you require Chinese output; default is English.
• For URLs, run DeepReader ingestion first so distill can target the correct notebook title.

Example use cases

• Convert a course NotebookLM notebook into an Obsidian Deep Q&A file for study.
• Run research --topic to gather web sources into a new NotebookLM notebook, then distill it into notes.
• Generate a 15–30 term glossary from a domain notebook and save as <NotebookName>_Glossary.md.
• Create ready-to-post JSON quizzes (quiz) and evaluate learner answers (evaluate) in Discord.
• Persist meeting summaries or AI-generated drafts directly into an Obsidian folder with YAML frontmatter.