NotebookLM Automation

Automate creation of multiple audio overviews in Google NotebookLM with different prompts and focus areas.

Installation

1. Download and extract this zip file to your server:

scp notebooklm-automation.zip your-server:~
ssh your-server
unzip notebooklm-automation.zip
cd notebooklm-automation

2. Install dependencies:

pip install playwright --break-system-packages
playwright install chromium

3. Make script executable:

chmod +x scripts/generate_audio_overviews.py

Quick Start

1. Create an input file (my_project.json):

{
  "sources": [
    "https://example.com/article1",
    "https://example.com/article2"
  ],
  "audio_overviews": [
    {
      "prompt": "Focus on technical concepts for developers",
      "length": "longer"
    },
    {
      "prompt": "Create executive summary for business leaders",
      "length": "default"
    }
  ]
}

2. Run the automation:

python scripts/generate_audio_overviews.py my_project.json

3. Audio files download to current directory

Usage Options

Basic (with visible browser for authentication)

python scripts/generate_audio_overviews.py input.json

Headless mode (after initial auth)

python scripts/generate_audio_overviews.py input.json --headless

Custom timeout (default is 10 minutes)

python scripts/generate_audio_overviews.py input.json --timeout 900

Input File Format

{
  "sources": [
    "URL1",
    "URL2"
  ],
  "audio_overviews": [
    {
      "prompt": "What the AI should focus on",
      "length": "default"  // or "longer" (never "shorter")
    }
  ]
}

See references/example_input.json for a complete example.

How It Works

The script uses Playwright to:

1. Open NotebookLM and create a new notebook
2. Add all source URLs
3. Wait for sources to process
4. Create each audio overview with its specific prompt/length
5. Wait ~10 minutes for generation
6. Download all audio files

For Your iOS + SSH Workflow

Since you operate from iOS via SSH:

1. Initial setup (one time):

ssh your-server
cd ~/notebooklm-automation
pip install playwright --break-system-packages
playwright install chromium

2. Each time you need audio overviews:

ssh your-server
cd ~/notebooklm-automation
vim input.json  # or upload via SFTP
python scripts/generate_audio_overviews.py input.json

3. Files download to /home/claude/ on your server

4. Retrieve files via SFTP or set up a simple web server

Authentication Handling

First run requires Google sign-in:

• Run without --headless flag
• Use X11 forwarding: ssh -X your-server
• Or use VNC to see the browser
• After first auth, subsequent runs can use --headless

Troubleshooting

See references/troubleshooting.md for:

• Updating selectors if NotebookLM UI changes
• Common failure scenarios
• Debugging tips

File Structure

notebooklm-automation/
├── README.md              # This file
├── SKILL.md              # Detailed skill documentation
├── INSTALL.md            # Installation guide
├── scripts/
│   └── generate_audio_overviews.py  # Main automation script
└── references/
    ├── example_input.json           # Example input file
    └── troubleshooting.md           # Troubleshooting guide

Limitations

• Max ~10 audio overviews per run
• Sources must be publicly accessible websites
• Requires Google account with NotebookLM access
• Generation takes ~5-10 minutes per overview
• UI automation may break if NotebookLM updates