Getting Started with Anam CLI

The anam binary (compiled from the anam-cli crate) is the main developer tool for starting the AnamDB server, initializing new workspace projects, checking health, and running the interactive REPL.

---

Installation

From Cargo (Rust Package Manager)

If you have Rust installed, you can build and install the binary from source:

cargo install --git https://github.com/AnamDB/anam-db anam-cli

Via Docker

If you prefer containerized execution, you can run the server directly:

docker run -p 8080:8080 ghcr.io/anamdb/anam-db

---

Workspace Initialization

To structure a local project, create a directory and initialize it with:

mkdir my-anam-project && cd my-anam-project
anam init

This scaffolds the following layout in your workspace:

├── anamdb.toml      # Project configuration
├── catalog.json     # Registers loaded tables, models, and rules
├── .env.example     # Environment template for keys
├── queries/
│   └── example.sql  # Boilerplate query structure
├── tables/          # Local Lance files
├── models/          # Local ONNX model files
└── packs/           # Logic pack configurations

---

Environment Configuration

Copy .env.example to .env to configure optional integrations like OpenAI for natural language translation:

# LLM API key for NL-to-Datalog compilation (optional)
ANAM_LLM_API_KEY=sk-your-openai-api-key
ANAM_LLM_ENDPOINT=https://api.openai.com/v1
ANAM_LLM_MODEL=gpt-4o

---

Workspace Workflow

Once your project directory is set up, you can:

1. Start the server using the workspace config:

   anam start

2. Connect or run the interactive shell:

   anam

3. Check the status of a running node:

   anam status