maestro-brainstorm

Orchestrate multi-role brainstorming into CSV-driven tasks and feature indexes.

511|63|Updated Mar 17, 2026
One-click install
npx skills add https://github.com/catlog22/Maestro-Flow --skill maestro-brainstorm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: maestro-brainstorm
Source: https://github.com/catlog22/Maestro-Flow/tree/main/.codex/skills/maestro-brainstorm
Command: npx skills add https://github.com/catlog22/Maestro-Flow --skill maestro-brainstorm

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates structured brainstorming by converting a raw topic into a CSV-driven, multi-wave pipeline that produces guidance, parallel role analyses, and a cross-role synthesis, removing manual coordination overhead and ensuring consistent context propagation across agents.

Core Features & Use Cases

  • Guidance-first decomposition: single-agent generation of a guidance-specification with terms, non-goals, and up to eight features that steer subsequent role analysis.
  • Parallel role analysis: spawn multiple role-specific agents (3–9) to analyze the topic from system, UI, product, testing, and domain perspectives with per-role artifact outputs.
  • Synthesis & feature index: single-agent synthesis that reconciles findings, resolves conflicts with explicit tags, and emits feature-specs and a feature-index.json.
  • CSV-as-source-of-truth: tasks.csv master state, per-wave CSVs, discoveries.ndjson append-only board, and clear error/timeout handling enable reproducible runs and resume/continue behavior.
  • Use Case: product and engineering teams use it to run rapid, architecture-aware ideation sessions and generate prioritized feature candidates and analysis artifacts.

Quick Start

Run maestro-brainstorm with your topic to generate guidance, parallel role analyses, and a synthesis feature index, for example: maestro-brainstorm "Build real-time collaboration platform"

Frequently Asked Questions about maestro-brainstorm

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

FAQPage Schema
How do I automate multi-role brainstorming to generate a feature index from a single topic?

Multi-role brainstorming is automated by orchestrating a CSV-driven pipeline that decomposes a topic into a guidance specification, spawns parallel role-specific agents for analysis, and synthesizes findings into a feature index. This removes manual coordination and ensures consistent context propagation.

How does guidance specification work for parallel role analysis in product ideation?

Guidance specification works by having a single agent generate terms, non-goals, and up to eight features from a topic. This specification steers subsequent parallel role analysis, ensuring spawned agents analyze the topic from system, UI, product, testing, and domain perspectives with consistent context.

Can I use CSV files to manage state and resume multi-agent ideation workflows?

Yes, you can use CSV files as the source of truth to manage state and resume multi-agent ideation workflows. A master tasks.csv state, per-wave CSVs, and an append-only discoveries.ndjson board enable reproducible runs, clear error handling, and continue behavior across spawned agents.

What is the best way to reconcile conflicts across parallel role analyses during synthesis?

The best way to reconcile conflicts across parallel role analyses is through single-agent synthesis. This process reconciles findings from spawned role-specific agents, resolves conflicts with explicit tags, and emits feature-specs alongside a consolidated feature-index.json artifact.

Do I need to spawn agents with controlled concurrency for structured brainstorming?

Spawning agents with controlled concurrency is required for structured brainstorming. The process orchestrates three to nine role-specific agents to analyze a topic from multiple perspectives, requiring controlled concurrency and context propagation to generate per-role artifacts without manual coordination.

What are the limitations of using a CSV master state for multi-wave brainstorming pipelines?

Limitations of using a CSV master state include the need to handle error and timeout scenarios explicitly during multi-wave pipelines. While per-wave CSVs and an append-only discoveries board support reproducible runs, the orchestration requires strict management of controlled concurrency and context propagation across spawned agents.