OpenHive
A zero-setup, model-agnostic runtime for multi-agent 'colonies' that coordinate through a shared ledger and a persistent plan, aimed at running production business processes reliably.
OpenHive is a zero-setup, model-agnostic runtime for colonies of agents: a persistent, client-facing lead agent called the Queen, plus however many worker agents a job needs. A person describes the outcome they want, and the Queen does the work and grows a colony of worker clones around it to complete that work reliably and at scale.
Every agent, Queen or worker, shares one execution primitive, an agent loop, so there is no workflow graph to compile and no orchestration boilerplate to write. The colony coordinates through a shared tracker ledger and a persistent task plan, with crash-safe state that can be paused and resumed, deep observability, cost enforcement, and out-of-band human-in-the-loop review built into that same primitive.
It also provides general compute and browser use through a native extension, and support for custom models. It is written in Python and licensed under Apache-2.0.