AI Agent Environment Setup

This project sets up an AI agent environment using agent-zero and ollama in Docker containers.

Setup Instructions

1. Clone this repository

2. Configure environment variables:

   • Edit the .env file in the root directory
   • Variables are grouped using # [header] format
   • Required sections:
     • [Local] - Instance configuration
     • [ollama] - Ollama model settings
     • [agent-zero] - Agent configuration

3. Build the environment:

   • Windows: Run .\build.bat
   • This will:
     • Copy environment variables to required locations
     • Build Docker images
     • Start containers using docker-compose

Environment Structure

The environment consists of two main components:

• Ollama: Runs the AI models
• Agent-Zero: Handles agent orchestration

Container Setup

The build script will call the copy_env script which in turn will make sure the containers have the necessary keys in a .env file in the folders containing the Dockerfile.

The script will continue with building the specified images using the Dockerfiles:

REM cd into the agent-zero directory and build the docker image with a specific name
docker build --no-cache -t agent-zero ./agent-zero

Finally, the script will run the containers using a compose file in daemon mode:

REM Run the docker-compose file
docker-compose up -d

Todo:

• reinforcement learning
• production server
• hosting