<div align="center">

<img src="assists/images/logo.png" alt="DeepDoc Logo" width="150"/>

DeepDoc: The Local AI-Powered Documentation Generator

</div>

---

DeepDoc is a powerful, local-first Command Line Interface (CLI) tool and FastAPI service designed to generate comprehensive, publication-ready Markdown documentation for any codebase. Leveraging the power of Ollama, LangChain, and LangGraph, DeepDoc ensures your project documentation is always accurate, detailed, and up-to-date, all while keeping your code secure on your local machine.

🚀 Key Features & SEO Highlights

DeepDoc is engineered for efficiency and security, offering a suite of features that make documentation a seamless part of your development workflow.

| Category | Feature | Description | SEO Keywords | | :--- | :--- | :--- | :--- | | Local-First AI | 🧠 Offline Documentation Generation | Utilizes Ollama for completely offline, on-premise documentation. Your code never leaves your machine. | Local AI, Offline Documentation, Ollama, Secure Documentation | | Deployment | 🐳 Docker & API Support | Includes Dockerfile and docker-compose.yml for easy containerization and deployment as a FastAPI service. | FastAPI, Docker, Containerized AI, Documentation API | | Code Analysis | 🔍 Intelligent Project Scanner | Recursively scans directories, respects .gitignore, and accurately detects programming languages, frameworks, and dependencies. | Codebase Analysis, Project Scanner, Git Ignore Support, Dependency Analysis | | Output Quality | 📚 Comprehensive Output Suite | Generates a full suite of documents: README.md, SUMMARY.md, ARCHITECTURE.md, API.md, and CONTRIBUTING.md. | Auto-generate README, API Reference Generation, Architecture Documentation | | Extensibility | ⚙️ Highly Customizable Workflow | Configure tone, verbosity, LLM model, and file filtering via a simple config.yaml. Orchestrated by a robust LangGraph workflow. | LangChain Workflow, LangGraph, Customizable LLM, YAML Configuration | | Developer Experience | 🎨 Rich CLI & Interactive Docs | Features a beautiful, rich terminal UI with progress bars and colorful output. Generated Markdown includes interactive elements like collapsible sections and Mermaid diagrams. | CLI Tool, Rich Terminal UI, Mermaid Diagrams, Interactive Markdown |

⚙️ Prerequisites

Before you can unleash the power of DeepDoc, ensure the following are installed and running on your system:

1. Python 3.12+: The core language environment. (Updated from 3.11)

   python3 --version
   # Must be 3.12 or higher
   

2. Ollama: The local LLM runtime.
   • Install from the official Ollama website.
   • Ensure the Ollama server is running: ollama serve
   • Pull a model for use (e.g., llama3.2 or codellama):

     ollama pull llama3.2
     

3. Docker (Optional): Required for running the containerized API service.

⬇️ Installation

For detailed, step-by-step installation instructions, please refer to the dedicated INSTALL.md file.

Quick Install (Recommended):

1. Clone the Repository:

   git clone https://github.com/Alqudimi/DeepDoc.git
   cd DeepDoc
   

2. Install Dependencies in Editable Mode:

   pip install -e .
   

   This method installs DeepDoc as a package, making the docgen command available globally.

Docker Deployment

To run DeepDoc as a containerized API service:

1. Build the Docker image:

   docker-compose build
   

2. Run the service:

   docker-compose up
   

   The API will be available at http://localhost:5000. Refer to the API_README.md for endpoint details.

💡 Usage Guide

DeepDoc provides a simple, intuitive command-line interface. For a comprehensive guide, see USAGE.md.

1. Initialize Configuration (Optional)

Generate a config.yaml file to customize the documentation process, including LLM model choice, style, and file filtering.

python main.py init
# A config.yaml file is created in your current directory

2. Generate Documentation

Run the generate command to analyze your project and produce the documentation files in the specified output directory (default: docs/).

| Command | Description | Example | | :--- | :--- | :--- | | Current Directory | Generates docs for the directory where the command is run. | python main.py generate | | Specific Path | Generates docs for a target project path. | python main.py generate /path/to/your/project | | Custom Config | Use a configuration file other than the default config.yaml. | python main.py generate -c custom-config.yaml | | Specify Model | Override the default Ollama model set in the config. | python main.py generate -m codellama | | Overwrite | Force overwrite of existing documentation files. | python main.py generate --overwrite |

🏗️ Project Structure & Generated Output

The project now includes both a CLI and a robust API component.

DeepDoc/
├── src/
│   ├── api/                   # 🐳 FastAPI service for documentation generation
│   │   ├── main.py            # API entry point (runs on Uvicorn)
│   │   ├── routes/            # API endpoints (analyze, config, docs, status)
│   │   └── services/          # Core documentation service logic
│   └── docgen/                # 💻 Core CLI logic
│       ├── core/              # Configuration, Scanner, LLM Client, LangGraph Workflow
│       ├── analyzers/         # Code and Dependency Analyzers
│       └── generators/        # Documentation Writers
├── tests/
│   └── api/                   # Unit and integration tests for the API
├── docs/                      # Default output directory for generated documentation
├── main.py                    # CLI entry point
├── pyproject.toml             # Project metadata and dependencies (Python 3.12+)
├── Dockerfile                 # Docker build file for the API service
├── docker-compose.yml         # Docker orchestration for easy setup
├── API_README.md              # Documentation for the API service
├── USAGE.md                   # Detailed usage guide for the CLI
└── README.md                  # This file

🛡️ Privacy and Security

DeepDoc is built with a strong commitment to privacy and security:

• 100% Local Processing: All code analysis and documentation generation occur exclusively on your local machine or within your private Docker container.
• No External APIs: Your codebase is never transmitted to any external service or cloud API.
• Offline Operation: Once the Ollama model is downloaded, the tool can run completely offline.

🤝 Contributing to DeepDoc

We welcome contributions from the community! Whether it's adding new features, improving the prompt engineering, or enhancing the code analysis capabilities, your input is valuable.

1. Fork the DeepDoc repository.
2. Clone your fork.
3. Create a new feature branch.
4. Make your changes and ensure tests pass.
5. Commit your work and push to your fork.
6. Open a Pull Request to the main branch of the original repository.

📄 License

This project is licensed under the MIT License.

🧑‍💻 Developer & Repository

| Detail | Value | | :--- | :--- | | Developer | Abdulaziz Alqudimi | | Email | eng7mi@gmail.com | | Repository | https://github.com/Alqudimi/DeepDoc |

---

<div align="center">

DeepDoc: Document Smarter, Not Harder. Built with ❤️ for the developer community.

</div>