prose

Orchestrate multi-agent AI workflows using structured .prose programs.

1|Updated May 3, 2026
One-click install
npx skills add https://github.com/brikkoAI/brikko-studio --skill prose-brikkoai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prose
Source: https://github.com/brikkoAI/brikko-studio/tree/main/packages/core/extensions/open-prose/skills/prose
Command: npx skills add https://github.com/brikkoAI/brikko-studio --skill prose-brikkoai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

OpenProse is a programming language for AI sessions that enables orchestrating multiple AI agents through structured prose programs. It solves the problem of coordinating complex multi-agent workflows, managing state across sessions, and executing deterministic AI-driven pipelines without manual intervention.

Core Features & Use Cases

  • Multi-Agent Orchestration: Define agents with specialized roles such as researchers, reviewers, and critics, then coordinate their execution through sessions, parallel blocks, and reusable workflow templates.
  • State Management: Persist execution state across sessions using filesystem, in-context, SQLite, or PostgreSQL backends, enabling long-running workflows and resumption after interruption.
  • Program Composition: Import and compose modular programs, pass context between sessions, and build complex pipelines using loops, conditionals, and error handling.
  • Use Case: A development team can use OpenProse to automate code review pipelines where a captain agent coordinates parallel security, performance, and style reviewers, then synthesizes findings into actionable feedback.

Quick Start

Use the prose skill to execute a .prose program that orchestrates multiple AI agents for parallel research, review, and synthesis tasks.

Frequently Asked Questions about prose

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

FAQPage Schema
How do I orchestrate multi-agent AI workflows for parallel code review?

Multi-agent AI workflows can be orchestrated using structured prose programs that define specialized agents, sessions, and control flow. This enables parallel execution where a captain agent coordinates security, performance, and style reviewers to synthesize actionable feedback.

What is the best way to manage state persistence across multiple AI sessions?

State persistence across AI sessions is managed through filesystem, in-context, SQLite, or PostgreSQL backends. This enables long-running workflows and allows execution to resume smoothly after an interruption.

Can I use this multi-agent orchestration with Claude Code or OpenCode?

Yes, multi-agent orchestration supports Claude Code, OpenCode, and Amp harnesses. It requires the OpenProse VM runtime to execute the prose programs that define the AI agents and control flow.

How do I build complex AI pipelines with loops and conditionals?

Complex AI pipelines are built using modular program composition that supports loops, conditionals, and error handling. You can import and compose modular programs while passing context between sessions.

Does parallel execution support recursive language model patterns?

Yes, parallel execution supports recursive language model patterns and research synthesis. It applies to complex automation scenarios requiring modular composition of AI tasks and deterministic pipeline execution without manual intervention.