AI Book Writer

A Python-based system that uses AutoGen to generate complete books through collaborative AI agents. The system employs multiple specialized agents working together to create coherent, structured narratives from initial prompts.

Features

• Multi-agent collaborative writing system
• Structured chapter generation with consistent formatting
• Maintains story continuity and character development
• Automated world-building and setting management
• Support for complex, multi-chapter narratives
• Built-in validation and error handling

Architecture

The system uses several specialized agents:

• Story Planner: Creates high-level story arcs and plot points
• World Builder: Establishes and maintains consistent settings
• Memory Keeper: Tracks continuity and context
• Writer: Generates the actual prose
• Editor: Reviews and improves content
• Outline Creator: Creates detailed chapter outlines

Installation

1. Clone the repository:

git clone https://https://github.com/psylsph/ai-book-writer.git
cd ai-book-writer

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Set your API Key

• Update config.py with your DeepSeek API Key

Usage

1. Basic usage: python main.py

Configuration

The system can be configured through config.py. Key configurations include:

• LLM endpoint URL
• LLM API Key

Output Structure

Generated content is saved in the book_output directory:

book_output/
├── outline.txt
├── chapter_01.txt
├── chapter_02.txt
└── ...

Requirements

• Python 3.8+
• AutoGen 0.2.0+
• Other dependencies listed in requirements.txt

Development

To contribute to the project:

1. Fork the repository
2. Create a new branch for your feature
3. Install development dependencies:

pip install -r requirements.txt

4. Make your changes
5. Run tests:

pytest

6. Submit a pull request

Error Handling

The system includes robust error handling:

• Validates chapter completeness
• Ensures proper formatting
• Maintains backup copies of generated content
• Implements retry logic for failed generations

Limitations

• Requires significant computational resources
• Generation time increases with chapter count
• Quality depends on the underlying LLM model
• May require manual review for final polish

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Built using the AutoGen framework
• Inspired by collaborative writing systems