NotebookLM Research Assistant

Query Google NotebookLM to get source-grounded answers from Gemini. Each question opens a browser session, retrieves the answer exclusively from uploaded documents, and closes.

Based on PleasePrompto/notebooklm-skill.

Core Principle

NotebookLM answers come ONLY from user-uploaded documents — zero hallucination by design. Use this as a grounded research layer before synthesizing your own response.

Critical Rules

1. ALWAYS use run.py wrapper — never call scripts directly
2. NEVER guess notebook metadata — use Smart Add or ask the user
3. ALWAYS follow up — when you see "Is that ALL you need to know?", analyze gaps before responding

# CORRECT:
python scripts/run.py ask_question.py --question "..."