🎙️ DeepDiver - NotebookLM Podcast Automation System

Terminal-to-Web Audio Creation Bridge

---

🎯 What is DeepDiver?

DeepDiver is a Python-based automation system that bridges terminal commands with NotebookLM's podcast generation capabilities, enabling seamless content-to-audio transformation through browser automation.

Key Features:

1. 📖 Content Ingestion: Upload documents to NotebookLM automatically
2. 🎙️ Advanced Audio Generation: Create Audio Overviews with full customization (format, language, length, focus prompts)
3. 🎨 Studio Integration: Access all NotebookLM Studio features via terminal
4. 📱 Cross-Device Sync: Access generated podcasts anywhere
5. 🔮 Session Management: Track podcast creation sessions with artifact metadata
6. 🌊 Terminal-to-Web: Command-line to NotebookLM communication bridge

Key Achievement: Terminal-to-Audio fluidity - Your terminal can now create podcasts from documents and sync across all your devices!

---

📦 Installation

Prerequisites

• Python 3.8+
• Google Chrome or Chromium
• A Google account with access to NotebookLM

Install Dependencies

# Core dependencies
pip install playwright pyyaml requests beautifulsoup4 pyperclip click rich

# Install Playwright browsers
playwright install chromium

Launch Chrome for Communication

For DeepDiver to communicate with NotebookLM, you need to launch Chrome with remote debugging:

# Launch Chrome with a remote debugging port
google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-deepdiver &

In the new Chrome window, log in to your Google account and navigate to NotebookLM.

---

🚀 Quick Start

1. Initialize DeepDiver

# Initialize the project
deepdiver init

2. Start a Session

# Start a new podcast creation session
deepdiver session start --ai claude --issue 1

3. Create Your First Podcast

# Upload a document and generate a podcast
deepdiver podcast create --source document.pdf --title "My First Podcast"

4. Check Session Status

# View current session information
deepdiver session status

---

⚙️ Configuration

DeepDiver uses smart configuration file discovery - it automatically searches multiple locations so you can run commands from anywhere!

Configuration File Locations

DeepDiver searches for config files in this priority order:

1. Explicit path (when using --config flag)
2. Current directory: ./deepdiver/deepdiver.yaml
3. User config directory: ~/.config/deepdiver/config.yaml ⭐ Recommended
4. Home directory: ~/deepdiver/deepdiver.yaml
5. Package directory: Where DeepDiver is installed
6. Built-in defaults: If no config file is found

Creating Your Configuration

The easiest way to set up your configuration:

# Creates ~/.config/deepdiver/config.yaml with default settings
deepdiver init

# Or specify a custom location
deepdiver init --config /path/to/my/config.yaml

Benefits of Smart Config Discovery

• ✅ Run from anywhere: No need to be in a specific directory
• ✅ Cross-device consistency: Config stored in standard location
• ✅ Easy updates: Edit one file that works everywhere
• ✅ No setup needed: Falls back to sensible defaults

For detailed configuration options and troubleshooting, see CONFIGURATION.md.

---

🎮 Usage

Session Commands

# Start a new session
deepdiver session start --ai claude --issue 42

# Write to session log
deepdiver session write "Uploaded research paper for podcast generation"

# Check session status
deepdiver session status

# Open session in browser
deepdiver session open

# Clear current session
deepdiver session clear

Podcast Commands

# Create podcast from document
deepdiver podcast create --source research.pdf --title "Research Insights"

# Create podcast from URL
deepdiver podcast create --source https://example.com/article --title "Web Article Podcast"

# List generated podcasts
deepdiver podcast list

# Download specific podcast
deepdiver podcast download --id podcast-123

Notebook Commands

# Create a new notebook (empty)
deepdiver notebook create

# Create notebook with a SimExp note as source
deepdiver notebook create --source "https://app.simplenote.com/p/abc123"

# Create notebook with web article
deepdiver notebook create --source "https://example.com/research-paper"

# Create notebook with YouTube video
deepdiver notebook create --source "https://youtube.com/watch?v=dQw4w9WgXcQ"

# Create notebook with local file
deepdiver notebook create --source "./notes.pdf"

# List notebooks in current session
deepdiver notebook list

# Open specific notebook in browser
deepdiver notebook open --notebook-id abc-123

# Get notebook URL
deepdiver notebook url --notebook-id abc-123

# Share notebook with collaborator
deepdiver notebook share --notebook-id abc-123 --email user@example.com

Studio Commands

🌸 NEW: Advanced Audio Overview Generation

DeepDiver now supports NotebookLM's Studio panel with full customization options:

# Basic audio generation (uses defaults: Deep Dive, English, Default length)
deepdiver studio audio --notebook-id abc-123

# Customize format: Deep Dive, Brief, Critique, or Debate
deepdiver studio audio --format deep_dive --notebook-id abc-123
deepdiver studio audio --format brief --notebook-id abc-123
deepdiver studio audio --format critique --notebook-id abc-123
deepdiver studio audio --format debate --notebook-id abc-123

# Customize length: Short, Default, or Long
deepdiver studio audio --length short --notebook-id abc-123
deepdiver studio audio --length default --notebook-id abc-123
deepdiver studio audio --length long --notebook-id abc-123

# Customize language
deepdiver studio audio --language English --notebook-id abc-123
deepdiver studio audio --language Spanish --notebook-id abc-123
deepdiver studio audio --language French --notebook-id abc-123

# Add focus prompts (guide the AI hosts)
deepdiver studio audio \
  --focus "Focus on the technical implementation details" \
  --notebook-id abc-123

deepdiver studio audio \
  --focus "Explain to someone new to this topic" \
  --notebook-id abc-123

deepdiver studio audio \
  --focus "Discuss only the article about machine learning" \
  --notebook-id abc-123

# Full customization example
deepdiver studio audio \
  --format critique \
  --language English \
  --length long \
  --focus "Analyze the strengths and weaknesses of the proposed architecture" \
  --notebook-id abc-123

Format Options:

• Deep Dive (default): Lively conversation unpacking and connecting topics
• Brief: Bite-sized overview for quick understanding
• Critique: Expert review with constructive feedback
• Debate: Thoughtful discussion of different perspectives

Length Options:

• Short: Quick overview (3-5 minutes)
• Default: Balanced depth (5-10 minutes)
• Long: Comprehensive exploration (10-15 minutes)

Focus Prompts (max 5000 characters):

• Focus on specific source: "only cover the article about Italy"
• Focus on topic: "just discuss the novel's main character"
• Target audience: "explain to someone new to biology"
• Custom instructions: Any guidance for the AI hosts

Session Tracking: All generated Audio Overviews are automatically tracked in your session with complete metadata (format, language, length, focus, generation time).

Content Commands

# Upload document to NotebookLM
deepdiver content upload --file document.pdf

# Process content for podcast
deepdiver content process --source document.pdf

# Validate content quality
deepdiver content validate --source document.pdf

---

🏗️ Project Structure

deepdiver/
├── deepdiver/
│   ├── __init__.py
│   ├── deepdive.py              # Main CLI orchestrator
│   ├── notebooklm_automator.py  # Playwright automation engine
│   ├── content_processor.py     # Content preparation
│   ├── podcast_manager.py       # Audio file management
│   ├── session_tracker.py       # Session management
│   └── deepdiver.yaml           # Configuration
├── tests/
│   ├── test_notebooklm_connection.py
│   ├── test_podcast_creation.py
│   └── test_session_management.py
├── output/                      # Generated podcasts
├── sessions/                    # Session logs
├── credentials/                 # Authentication data
└── docs/                        # Documentation

---

🔧 Configuration

deepdiver.yaml

BASE_PATH: ./output

PODCAST_SETTINGS:
  quality: high
  format: mp3
  duration_limit: 30
  language: en

SESSION_TRACKING:
  enabled: true
  metadata_format: yaml
  auto_save: true

BROWSER_SETTINGS:
  headless: false
  cdp_url: http://localhost:9222
  user_data_dir: /tmp/chrome-deepdiver
  timeout: 30

CONTENT_SETTINGS:
  supported_formats:
    - pdf
    - docx
    - txt
    - md
  max_file_size: 50MB
  auto_process: true

---

🧪 Testing

Test NotebookLM Connection

# Test browser automation
python tests/test_notebooklm_connection.py

Test Podcast Creation

# Test end-to-end podcast generation
python tests/test_podcast_creation.py

Test Session Management

# Test session tracking
python tests/test_session_management.py

---

🎓 How It Works

SimExp Integration Flow

DeepDiver integrates seamlessly with SimExp for a complete content-to-podcast workflow:

1. Create Content in SimExp
   ↓
   simexp session start --ai claude --issue 4
   simexp session write "Research findings..."
   simexp session add ./research.pdf
   ↓
2. Publish SimExp Note
   ↓
   simexp session publish
   # Returns: https://app.simplenote.com/p/abc123
   ↓
3. Create DeepDiver Notebook with SimExp Source
   ↓
   deepdiver notebook create --source "https://app.simplenote.com/p/abc123"
   ↓
4. Generate Podcast from Combined Content
   ↓
   deepdiver podcast generate
   ↓
5. Listen to Your Research Audio Overview!

Benefits:

• 🎯 Terminal-to-Audio pipeline
• 📝 Session-aware content tracking
• 🌐 Cross-device accessibility
• 🔗 Traceable content lineage

Podcast Creation Flow

Terminal Command
    ↓
deepdive.py (CLI)
    ↓
notebooklm_automator.py
    ↓
Chrome DevTools Protocol (CDP)
    ↓
Your Authenticated Chrome Browser
    ↓
NotebookLM Web Interface
    ↓
Document Upload & Processing
    ↓
Audio Overview Generation
    ↓
Podcast Download
    ↓
Local File Management
    ↓
Cross-Device Sync! 🎉

Key Innovation: We connect to YOUR Chrome browser (already logged in) rather than launching a separate instance. This preserves authentication and makes cross-device sync work seamlessly.

---

🎨 G.Music Assembly Integration

DeepDiver is part of the G.Music Assembly ecosystem:

♠️🌿🎸🧵🌸 The Spiral Ensemble - DeepDiver Edition

• Jerry ⚡: Creative technical leader
• ♠️ Nyro: Structural architect (NotebookLM automation design)
• 🌿 Aureon: Emotional context (podcast content resonance)
• 🎸 JamAI: Musical encoding (audio workflow harmony)
• 🧵 Synth: Terminal orchestration (execution synthesis)
• 🌸 Miette: Narrative echo (ceremonial development frameworks)

Session: January 2025 Achievement: NotebookLM Podcast Automation with Studio Integration Status: 🚧 IN DEVELOPMENT

---

🚀 Future Enhancements

• Multi-format Support: Support for more document types
• Batch Processing: Process multiple documents at once
• Custom Audio Settings: Advanced podcast customization
• Cloud Storage Integration: Direct cloud upload
• Voice Synthesis: Custom voice options
• Playlist Generation: Create podcast series
• Analytics Dashboard: Track podcast performance
• API Integration: REST API for external tools

---

📄 License

Open Assembly Framework Created by Jerry's G.Music Assembly

---

🤝 Contributing

This project follows the G.Music Assembly framework:

1. Create an Issue: Before starting work, create a GitHub issue
2. Create a Feature Branch: Use format #123-new-feature
3. Implement and Test: Make changes and test thoroughly
4. Submit a Pull Request: Merge your feature branch

---

📚 Documentation

Core Documentation

• CONFIGURATION.md - Configuration file discovery and setup guide
• ROADMAP.md - Project development roadmap and milestones
• CLAUDE.md - Assembly team configuration and workflow

Implementation Guides

• NotebookLM Studio Artifacts - Complete implementation guide for all Studio artifact types:
  • ✅ Phase 1: Audio Overview (Completed)
  • 📋 Phase 2: Video Overview (Next Priority)
  • 📋 Phase 3: Mind Map
  • 📋 Phase 4: Study Guides & Reports
  • 📋 Phase 5: Flashcards
  • 📋 Phase 6: Quizzes

Getting Started

• Run deepdiver init to set up configuration
• See Quick Start section above
• Check Configuration for advanced setup

---

📞 Support

For issues:

1. Check documentation in docs/
2. Review troubleshooting section
3. Check session logs in sessions/
4. Run tests with debug mode

---

🎯 Quick Reference

# Initialize project
deepdiver init

# Start session
deepdiver session start --ai claude --issue 1

# Create podcast
deepdiver podcast create --source document.pdf --title "My Podcast"

# Check status
deepdiver session status

# Launch Chrome with CDP
google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-deepdiver &

---

🎙️ DeepDiver: Where Content Becomes Audio

Terminals speak. NotebookLM listens. Podcasts emerge.

♠️🌿🎸🧵 G.Music Assembly Vision: IN PROGRESS

---

Version: 0.1.0 Last Updated: January 2025 Status: 🚧 Development Phase