| name | notebooklm-skill |
| description | Orchestrates deep research and synthesis across complex document libraries (PDFs, docs, web). Use for multi-source analysis, grounded report generation, cross-referencing archives, or creating study aids from large document sets. Handles project discovery, source ingestion, and grounded reasoning for comprehensive research tasks. |
| license | Apache-2.0 |
| metadata | {"version":"1.0.0","language":"en","trit":0,"author":"agent-builder"} |
NotebookLM Research Assistant
Interact with Google NotebookLM to query your documentation library.
🚀 Self-Deployment & Bootstrapping
If this skill's research workflows (like /notebook-research, /notebook-study) are not appearing in your agent's slash-commands, run the following command from the skill's root directory:
python scripts/run.py bootstrap.py --workspace .
This will automatically detect your agent's configuration directory (e.g., .agents, .cursor, .gemini, or .agent) and deploy the necessary .md or .mdc files.
Essential Commands
ALWAYS use scripts/run.py wrapper.
1. Unified Interface (Recommendation)
Use the new bridge for all operations. It handles authentication automatically from storage.
python scripts/run.py unified_bridge.py list
python scripts/run.py unified_bridge.py create --title "Project Beta"
python scripts/run.py unified_bridge.py ask --notebook-id [ID] --question "..."
2. Research & Learning Workflows
Invoke these from the workspace root for high-level orchestration:
/notebook-research [TOPIC]: Deep dive research and source aggregation./notebook-study [ID]: Active learning with quizes/flashcards./notebook-visualize [ID]: Generate Mind-Maps.
3. Artifact Generation
python scripts/run.py unified_bridge.py artifact --notebook-id [ID] --type audio
python scripts/run.py unified_bridge.py artifact --notebook-id [ID] --type quiz
python scripts/run.py unified_bridge.py artifact --notebook-id [ID] --type mind-map
Workflow
- Discover / Sync: ALWAYS run
python scripts/run.py unified_bridge.py list first to see existing NotebookLM projects. Do not create new ones blindly; be aware of what is already present!
- Research: Call
/notebook-research to initialize.
- Optimize: Use
/gh-learn to find secondary sources and add them.
- Study: Run
/notebook-study to verify comprehension.
- Visualize: Use
/notebook-visualize for architectural insight.
References
