prose

Multi-agent workflow orchestration is achieved by defining execution flows using the declarative OpenProse programming language, which coordinates agents through a VM-like execution model for sequential, parallel, or conditional tasks.

Conditional execution and error handling for LLM agents are supported natively by the declarative programming language, allowing you to manage agent state and route execution dynamically without manual prompt engineering.

Automating a multi-step code review process involves defining multiple AI agents where one identifies bugs, another checks style compliance, and a third synthesizes feedback into a single report using structured workflow automation.

Executing declarative agent programs requires a Prose Complete system to interpret and run the OpenProse language through its VM-like execution model, managing state and agent coordination.

Getting started with declarative AI agent workflows involves running the provided example program, 'examples/01-hello-world.prose', to understand how the structured programming language defines and executes agents.

Declarative LLM orchestration eliminates manual agent coordination by defining agents with specific models, prompts, and skills in a structured program, ensuring reliable state management and complex execution flows.