Overview
Investigate NotebookLM MCP for integration into shipkit.
What it is: MCP server that connects Claude Code to Google's NotebookLM for advanced research workflows.
Context from Article
NotebookLM is Google's AI research tool that:
- Synthesizes information from multiple sources
- Creates audio overviews (podcast-style summaries)
- Organizes research notes
- Generates insights from documents
MCP integration enables:
- Using NotebookLM from within Claude Code
- Enhanced /research skill capabilities
- Multi-source research synthesis
- Audio summaries of research findings
Questions to Answer
1. Capabilities
- What can NotebookLM MCP do?
- Create notebooks programmatically?
- Query existing notebooks?
- Generate audio summaries?
- Export research findings?
2. Setup Requirements
- API Key: Google Cloud? NotebookLM-specific?
- Prerequisites: Node.js? Google account?
- Free tier: Available? Usage limits?
- Complexity: Easy setup or involved?
3. Use Cases
Who benefits:
- Researchers compiling information
- Developers investigating new technologies
- Technical writers gathering sources
- Anyone doing deep research (/research skill users)
When useful:
- Market/competitive research
- Technology evaluation
- Documentation deep-dives
- Learning new domains
Frequency:
- Daily: For research-heavy roles
- Weekly: General development research
- Rarely: Pure implementation work
4. Integration with Existing Skills
Enhances /research skill:
Current /research:
- Multi-source web search
- Citation management
- Confidence scoring
With NotebookLM:
- Persistent research notebooks
- Audio summaries of findings
- Better source organization
- Cross-session research continuation
Example workflow:
/research "serverless architecture patterns"
→ Compiles findings into NotebookLM
→ "Generate audio summary of research"
→ Listen to 10-min podcast while commuting
5. Where Does It Belong?
Option A: Core
✅ Research is universal (every dev researches)
❌ Requires Google account + API key
❌ Setup friction for casual users
❌ Not everyone needs audio summaries
Option B: Experimental
✅ New integration, unproven in shipkit
✅ Users can try it out
❓ Might be mature enough for Advanced
Option C: Advanced
✅ Specialized for research-heavy work
✅ Opt-in for those who need it
✅ Matches other domain-specific tools
✅ Likely best fit
Option D: Essential MCP (offered during install)
If NotebookLM is free and easy to set up, could be offered alongside Brave Search as research enhancement:
Essential (free, enhance core skills):
[ ] Brave Search - Web search for /research
[ ] NotebookLM - Advanced research synthesis
6. Integration Strategy
Add to /research skill:
- Check if NotebookLM MCP available
- If yes: "Save findings to NotebookLM notebook?"
- If no: "💡 Research would be enhanced with NotebookLM MCP. Install?"
Documentation:
- Add to docs/mcp-servers.md under "Research & Learning"
- Add to seed/mcp.sample.json
- Document API key setup
Potential wrapper skill:
/notebook-research - Research workflow using NotebookLM/audio-summary - Generate podcast summary of topic
7. Comparison to Existing
| Feature |
Current /research |
With NotebookLM |
| Web search |
✅ Via Brave MCP |
✅ |
| Citations |
✅ Built-in |
✅ Enhanced |
| Source synthesis |
✅ Basic |
✅✅ Advanced |
| Persistence |
❌ Per-session |
✅ Notebooks |
| Audio summaries |
❌ |
✅ Podcasts |
| Cross-session |
❌ |
✅ Resume research |
Significant enhancement to research capabilities.
Investigation Tasks
Recommendation Needed
After investigation:
Placement: [Core / Experimental / Advanced / Essential MCP]
Rationale:
- [Setup complexity]
- [Use case frequency]
- [Value vs friction]
Integration:
- [Enhance /research skill? Y/N]
- [Create wrapper skills? Y/N]
- [Offer during install? Y/N]
Related
All research/creative tools being investigated:
Pattern: Google AI tools + MCP integration = enhanced capabilities
Worth bundling as "Google AI Tools" package?
Overview
Investigate NotebookLM MCP for integration into shipkit.
What it is: MCP server that connects Claude Code to Google's NotebookLM for advanced research workflows.
Context from Article
NotebookLM is Google's AI research tool that:
MCP integration enables:
Questions to Answer
1. Capabilities
2. Setup Requirements
3. Use Cases
Who benefits:
When useful:
Frequency:
4. Integration with Existing Skills
Enhances /research skill:
Current /research:
With NotebookLM:
Example workflow:
5. Where Does It Belong?
Option A: Core
✅ Research is universal (every dev researches)
❌ Requires Google account + API key
❌ Setup friction for casual users
❌ Not everyone needs audio summaries
Option B: Experimental
✅ New integration, unproven in shipkit
✅ Users can try it out
❓ Might be mature enough for Advanced
Option C: Advanced
✅ Specialized for research-heavy work
✅ Opt-in for those who need it
✅ Matches other domain-specific tools
✅ Likely best fit
Option D: Essential MCP (offered during install)
If NotebookLM is free and easy to set up, could be offered alongside Brave Search as research enhancement:
6. Integration Strategy
Add to /research skill:
Documentation:
Potential wrapper skill:
/notebook-research- Research workflow using NotebookLM/audio-summary- Generate podcast summary of topic7. Comparison to Existing
Significant enhancement to research capabilities.
Investigation Tasks
Recommendation Needed
After investigation:
Related
All research/creative tools being investigated:
Pattern: Google AI tools + MCP integration = enhanced capabilities
Worth bundling as "Google AI Tools" package?