What problem does it solve? Coordinating multiple AI agents for complex tasks requires ad-hoc prompting and manual handoffs. OpenProse provides a structured programming language for AI sessions, letting you write reusable .prose programs that spawn subagents, manage state, handle errors, and run tasks in parallel. ## Core Features & Use Cases - Program Execution: Run local .prose files or fetch programs from URLs and the p.prose.md registry, with the AI embodying the OpenProse VM to spawn real subagent sessions. - Compilation and Validation: Validate .prose syntax against the full language grammar without executing, catching undefined agents, invalid models, and malformed constructs. - State Management: Choose between filesystem, in-context, SQLite, or PostgreSQL state backends, with persistent agents that retain memory across runs. - Use Case: Write a program that runs three code reviewers in parallel, loops until no bugs remain, and escalates based on AI-evaluated severity—then execute it with a single prose run command. ## Quick Start Run the hello world example by asking the assistant to execute prose run examples/01-hello-world.prose.
