prose

You orchestrate multi-agent AI workflows using a unified VM that centralizes agent definitions, sessions, and imports while managing persistent state across nested tasks. This structured state management ensures reliable execution traces for complex agent coordination.

The best way to coordinate nested AI agent sessions is using a unified runtime that supports nested imports and persistent agents. This approach automates complex multi-agent tasks and provides structured state backends for resumption and auditing.

Yes, you can use SQLite or PostgreSQL for multi-agent workflow state management. The system supports flexible backends including filesystem, SQLite, and PostgreSQL to suit different scales and ensure reliable execution traces across projects.

To set up a basic multi-agent workflow pipeline, define a basic agent and a simple session, then run a small multi-agent workflow. This quick start approach demonstrates the orchestration runtime in action with minimal manual coordination overhead.

Yes, multi-agent orchestration supports execution traces for auditing. By providing structured state management and flexible backends, the system satisfies requirements for reliable execution traces and scalable workflow pipelines across teams.