prose

You orchestrate multi-agent AI workflows by defining structured programs using the OpenProse language, which manages parallel execution, conditional logic, and state persistence. This turns complex agent sessions into predictable, repeatable processes.

Agent orchestration coordinates multiple AI agents with specific roles and models to execute chained tasks. Structured programming is needed when automating complex processes like code reviews, ensuring predictable state persistence and dynamic context passing.

Yes, you can run parallel AI agent tasks and pass dynamic context. The declarative syntax supports agent specialization, dynamic context passing, and recursive program composition via imports for complex workflow automation.

Start automating a code review by running the example program 'examples/01-hello-world.prose' using the prose skill. You can define one agent to identify bugs, another for style compliance, and a third to synthesize feedback.

Yes, this workflow automation approach handles error handling and state persistence. The structured, declarative syntax ensures multi-agent workflows maintain reliable state and manage execution errors predictably.

The best way to define roles and models is through a declarative syntax that supports agent specialization. This structured programming language allows you to assign specific models and skills to individual agents within the workflow.