Quick Start

What you need

• Python 3.10 or newer.
• A model. Either an API key for a hosted provider, or a local model server such as llama.cpp, LM Studio or Ollama.

Install and start

Download the latest build from Get AWORG, then:

pip install aworg-<version>-py3-none-any.whl
aworg install
aworg start

aworg install creates your Aworg's home, ~/.aworg, and seeds it with the personas AWORG ships. aworg start serves the interface on port 8420 and prints the address.

The first time it starts, AWORG generates an interface password and prints it. Sign in with that. You can change it later in Settings.

Connect a model

Open Settings → Connections and add your provider, or the address of your local model server. Test the connection, then choose it for the Resident.

Give it something to do

Say what you want in the chat. The Resident plans the work, does it in the workspace, and tells you what it built. The preview shows the result.

Teach it more

Browse the Skills Library and Tool Center, then install what you need:

aworg get skills/house-style
aworg get tools/weather
aworg get personas/wizard

A bare name works too (aworg get weather) when only one type has that name. A running Aworg picks up whatever you install without a restart.

Running it on a server

aworg start --host 0.0.0.0 makes it reachable from other machines. The interface password is what stands between the network and a Resident with a shell, so pick a strong one, and put AWORG only on networks you trust.