world-runtime gives domain-specific agents durable state, policy boundaries, provenance, and operator control on top of existing systems of record.
The missing layer is the runtime that lets agents operate safely between them. world-runtime coordinates state, events, policies, approvals, and agent behavior across enterprise workflows.
Track entities, events, and decisions over time instead of treating every AI interaction as an isolated request.
Keep human review, rule evaluation, and tool boundaries inside the runtime itself by design.
Build multiple domain-specific agent systems on one shared runtime instead of rebuilding the same trust layer for every workflow.
Why
Why reframes the problem: enterprise systems already hold truth and models already reason, but without governed state, policy, approvals, and intervention, agent systems drift.
Build
Build turns the runtime into concrete domains: supply networks, lab science, air traffic, power grids, city operations, and multi-agent systems, each proving a different operating constraint.
How
How focuses on the build pattern: keep your models, tools, services, and cloud choices, then use world-runtime to structure explicit state, proposals, simulation, and policy where consequential decisions need discipline.
Docs
Docs focuses the repo into usable paths: evaluator orientation, developer quickstarts, public APIs, operator playbooks, and extension contracts for teams that want to build seriously.
Repository
Explore world-runtime
Further reading
Robb Bush lays out the business-strategy logic behind world-runtime: why coordination, trust, governance, and orchestration-native operating models matter more as intelligence becomes abundant.