Axon Backend

The @axon-ai/backend package serves as the core data processing and storage component for the Axon developer tool. It receives real-time tracing data from your LangChain applications (via @axon-ai/langchain-tracer), stores it in a local SQLite database, and serves this data to the @axon-ai/dashboard for visualization.

This backend is designed to run locally alongside your development environment, providing a private and efficient way to monitor and debug your AI agents and chains without sending sensitive data to external services.

✨ Features

• Real-time Data Ingestion: Receives trace events from LangChain applications instantly.
• Local SQLite Storage: Persists tracing data locally for historical analysis and quick retrieval.
• WebSocket API: Communicates with the Axon Dashboard via WebSockets for live updates.
• REST API: Provides endpoints for the dashboard to fetch historical trace data.
• Seamless CLI Integration: Automatically managed by the @axon-ai/cli for an effortless developer experience.

🚀 Getting Started (via Axon CLI)

The Axon Backend is an integral part of the Axon developer tool and is typically managed by the @axon-ai/cli. You usually won't need to interact with this package directly.

1. Install the Axon CLI

If you haven't already, install the Axon CLI globally:

npm install -g @axon-ai/cli

2. Initialize Your Project

Navigate to your LangChain project directory and initialize Axon:

axon-ai init --project my-langchain-app

This command sets up the necessary configuration files and automatically initializes the database for the backend.

3. Start the Axon Services (Backend & Dashboard)

From your project directory, start both the Axon backend and dashboard:

axon-ai start

This command will launch the backend server (usually on http://localhost:3000) and the dashboard, connecting them automatically.

🛠️ Direct Usage (for Development/Debugging)

If you need to run the backend independently for development or debugging purposes:

1. Clone the Axon repository:

   git clone https://github.com/your-axon-repo/Axon.git
   cd Axon/backend

2. Install dependencies:

   npm install

3. Initialize the database schema:

   npm run db:init

4. Start the backend in development mode (with auto-reloading):

   npm run dev

5. Start the backend in production mode:

   npm run build
   npm start

⚙️ Configuration

The backend's port and host can be configured via the .axon-ai/config.json file in your project root (created by axon-ai init) or by passing options to the axon-ai start command.

Example config.json snippet:

{
  "backend": {
    "port": 3000,
    "host": "localhost"
  }
}

🗄️ Database

The backend uses SQLite for local data persistence.

• Schema Initialization: The npm run db:init script creates the necessary tables. This is automatically handled by axon-ai init.
• Data Storage: All trace events, including LLM calls, tool usage, and chain executions, are stored in this local database.

🐛 Troubleshooting

If you encounter issues with the backend:

• Check CLI Status: Use axon-ai status to see if the backend server is running.
• Port Conflicts: If the backend fails to start, another process might be using port 3000. You can specify a different port using axon-ai start -p <new-port>.
• Logs: Check the console output where the backend is running for any error messages.

🤝 Contributing

Contributions are welcome! If you're interested in improving the Axon Backend, please see the main Axon repository for contribution guidelines.

📄 License

This project is licensed under the MIT License.