Core Concepts
Understand the architecture that makes Orrerygen unique.
Agents & Teams
Agents are autonomous units of execution. In Orrerygen, every agent runs in its own isolated Docker container. This ensures that dependencies (Python versions, system libraries) never conflict.
Single Agent
A customized LLM wrapper with specific tools (e.g., "Web Scraper", "Code Writer").
Team
A graph of agents collaborating. Example: A "Manager" agent delegating tasks to a "Researcher" and a "Writer".
Local Sovereignty
Cloud platforms require you to upload your data and API keys. Orrerygen is different. It follows a "Code to Data" model. We send the agent code to your machine, where it executes locally.
Local Persistence
All agent memory, logs, and outputs are stored in your local filesystem.
Secure Networking
Arbitrary internet access is blocked by default. Agents only communicate with the Orrerygen platform for license verification and authorized API usage.
Model Inference
Orrerygen is model-agnostic. You can connect agents to: