prose

Orchestrate multi-agent workflows in the OpenProse VM with sessions and imports.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/lalomorales22/penguins --skill prose-lalomorales22
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prose
Source: https://github.com/lalomorales22/penguins/tree/main/extensions/open-prose/skills/prose
Command: npx skills add https://github.com/lalomorales22/penguins --skill prose-lalomorales22

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

OpenProse provides a centralized, self-hosted orchestration layer for AI agent workflows. It enables teams to model, coordinate, and audit complex multi-agent pipelines within the OpenProse VM, from simple sessions to nested imports and parallel executions.

Core Features & Use Cases

  • Spawns and coordinates sessions (agents) and imported programs
  • Supports parallel, loops, pipelines, error handling, and context passing
  • Durable state backends (filesystem, in-context, SQLite/PostgreSQL) for long-running workflows
  • Reusable blocks and block-level scoping for modular design

Quick Start

Create a simple OpenProse workflow by defining sessions and a top-level skill pack, then run prose to observe outputs and bindings.

Frequently Asked Questions about prose

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate multi-agent workflows with parallel execution?

Multi-agent workflows with parallel execution are coordinated by spawning sessions inside the OpenProse VM, where agents run concurrent tasks with traceable bindings and shared context. The system supports loops and pipelines to manage complex concurrent operations.

Can I persist state for long-running multi-agent pipelines using SQLite or PostgreSQL?

Yes, long-running multi-agent pipelines can persist state using SQLite, PostgreSQL, filesystem, or in-context backends. This durability ensures that session history and traceable bindings remain intact across extended workflow executions.

How do I build conversational AI pipelines with reusable blocks and error handling?

Conversational AI pipelines are built by defining reusable blocks with block-level scoping, then using try/catch constructs for error handling. The OpenProse VM enforces clear contracts for inputs, outputs, and context passing between sessions.

Does this orchestration tool require any external dependencies to run agent sessions?

No external dependencies are required to run agent sessions. The orchestration layer is self-hosted within the OpenProse VM, allowing teams to model, coordinate, and audit multi-agent pipelines without relying on external service integrations.

What is the best way to audit and trace context passing in nested agent imports?

Auditing nested agent imports is handled by the OpenProse VM's enforced contracts for inputs, outputs, and context. It ensures traceable bindings and maintains history across spawned sessions and imported programs throughout the workflow.