Operon Documentation

Biologically inspired architectures for more reliable AI agent systems.

Operon is a research-grade Python library for building agent systems with explicit control structure. The current recommended entry point is the pattern-first API, with the deeper wiring and theory layers still available underneath.

Start Here

Getting Started Install the library and start from the smallest practical entry path. Pattern-First API Topology advice, reviewer gates, specialist swarms, and skill organisms. Skill Organisms Multi-stage workflows with fast/deep model routing and attachable components. Bi-Temporal Memory Append-only factual memory with dual time axes for auditable decision-making. Examples Where to start in the example catalog without reading everything in order.

Deeper Material

Concepts and Architecture The library’s current layers and the role of structure. Theory and Papers The whitepaper, the epistemic topology paper, and how to read them. API Overview A lightweight map of the namespaces and their intended use. Hugging Face Spaces Interactive demos for topology and diagram exploration. Release Notes Recent direction and public writing around v0.17 through v0.19.

The pattern-first API is now the public front door:

advise_topology(...), reviewer_gate(...), specialist_swarm(...), and skill_organism(...).

Docs source lives in the Operon repo under docs/site/.