convez.ai
AgenteProdutividadeIntermediário
sandhuka

Muster — Time de IA para Gestão de Produtos

sandhuka/muster-ai

Equipe de IA especializada com memória persistente para construir produtos completos

Ver no GitHub
InicialPluginScore · de 0 a 100

Framework multi-agente que transforma Claude Code em um time coordenado de oito especialistas (PM, dev, design, QA, conteúdo, marketing, legal, pesquisa) com memória persistente e contexto filtrado. Elimina amnésia entre sessões armazenando decisões e tarefas em arquivos markdown, permitindo que cada agente mantenha continuidade sem re-briefings.

Instalação

$npx convez-ai sandhuka/muster-ai

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
5
Forks
2
Acompanhando
0
Saúde
42%
Último commit
há 3 dias
Versão
v4.2
Issues abertas
1
Contribuidores
1

Casos de uso

  • Coordenar design, desenvolvimento e QA simultaneamente em um único projeto
  • Manter contexto de produto consistente entre múltiplas sessões Claude
  • Acelerar descoberta de produto e planejamento de sprint com especialistas IA

Veredito

Abordagem pragmática para orquestração multi-agente que prioriza persistência e contexto filtrado sobre comunicação complexa. Validado em projeto real; excelente para founders e pequenos times que precisam cobrir múltiplos domínios sem infraestrutura de framework.

README

A tradução para o português deste README ainda está na fila.

Muster

Ship a product. Without a team.

Eight AI specialists. Persistent memory. Just Claude Code.

Muster sprint status — PM coordinating multiple agents across a real iOS project

A real product needs design, code, QA, content, marketing, legal, and research.

You can't hire all of that. You don't have to.

Muster turns Claude Code into a coordinated team of eight AI specialists with persistent memory, quality guardrails, and conversational continuity. Every decision persists. Every sprint plans itself forward. Every specialist remembers the discussion across follow-ups — no re-briefing on every turn, no Claude amnesia between sessions.

Just markdown files. No external frameworks. No API wiring. No subscriptions.

You (Founder)
     |
     v
  Open Claude in project   →   Pick a role at session start
                                       |
                                       v
                              Session bound to ONE role:
                       PM | Dev | UI/UX | QA | Content | Mkt | Legal | Research

The PM coordinates: plans sprints, makes decisions, cascades context to specialists. Specialists do the domain work in their own session and file handoffs. Status line shows [muster: <role>] so you always know which tab is which.

Built and validated on a real iOS app, mid-construction. Not framework theory.

Just open Claude Code. Pick a role. Ship.

The problem

Most multi-agent frameworks optimize for agent communication — how agents pass messages to each other. The real bottleneck is context. Claude Code agents forget everything between sessions. If you're building a product across design, dev, legal, marketing, and QA, you need persistent memory and a way to keep each agent focused on what matters.

How Muster solves it

Three-tier reading model

Each agent reads roughly 80 lines at startup — its role, filtered product context, and the current task. Full docs are loaded on demand. Files meant for other agents are never loaded at all.

PM as context translator

When a decision is made, the PM doesn't tell every agent to go read the decision log. The PM updates each agent's context file with only what that agent needs. Decisions cascade through filtered summaries, not shared inboxes.

Files as persistent memory

Agent brains, orchestration queues, handoff logs, and decision records persist as markdown. No session starts from zero — every agent picks up where the last one left off.

Self-improving skills

During sprint planning, the PM scans for methodology gaps. New skills are classified as generic or product-specific. Generic skills are contributed back to the framework, so every project makes Muster sharper for the next one.

Quick start

New project

curl -fsSL https://raw.githubusercontent.com/sandhuka/muster-ai/main/scripts/setup-project.sh | bash -s my-project
cd ~/Desktop/my-project
claude

When Claude starts in a fresh project, the greenfield welcome fires automatically (no picker — onboarding is PM-driven). Tell Claude your product idea. The welcome walks you through five Discovery stages — idea share, market research, go/no-go decision, draft review, Sprint 1 plan. Total founder time is ~1–2 hours across ~3 sessions over a day or two.

See getting-started.md for the full walkthrough.

Existing project

If you already have code — a mobile app you've been building for months, a web app with real users, a backend service, a CLI tool — Muster has a dedicated adoption path. It reverse-discovers your product into a populated knowledge base and plans Sprint 1. Takes about two hours of focused time.

cd ~/path/to/your-existing-project
curl -fsSL https://raw.githubusercontent.com/sandhuka/muster-ai/main/scripts/setup-existing-project.sh | bash

See adopting-existing-project.md for the full walkthrough.

How it works

You describe your idea in the first session — PM is auto-bound for the greenfield welcome. PM sends Research to investigate the market, then writes the product spec, plans a sprint, and queues up agent tasks if the idea is viable.

After Discovery, you follow the orchestration queue. For each step, open a Claude session — pick the listed role from the picker (or set MUSTER_ROLE=<role> to skip the picker for scripts/automation). Sessions can run in parallel across separate terminals when tasks are independent. Each session reads its filtered context, does the work, files a handoff, and promotes the next step. Repeat until shipped.

The status line shows [muster: <role>] so you always know which role this session is bound to. /rebind swaps roles mid-session if you picked wrong.

Three ways to run it

Muster runs the same team three ways — pick by how much you want to drive.

ModeWhat it isUse it when
ManualA warm tab per role; you hop between themYou're thinking with one agent — planning, debugging, design
AssistedOne PM tab spawns each specialist as a subagentPM coordinates a short sequence while you watch
AutonomousA script walks the queue unattended in a git worktreeYou've planned a sprint and want to walk away

Planning is always interactive; execution is where you choose — and you can switch mid-sprint. See operating-modes.md for the full guide, including the cost and quality tradeoffs.

Framework help — the Guide

Stuck on how Muster runs — which mode to use, why a run stopped, tuning a budget, upgrading? Type /muster in any tab. The Guide is Muster's built-in concierge for the framework itself: it answers process questions — operating modes, run status, .muster/config knobs, version upgrades, friction reports — and routes product questions (specs, decisions, sprint content) to PM. It rides alongside your current role: in a role-bound tab, /muster answers once without changing what that tab is bound to.

Agent roster

AgentRole
PMPlans sprints, cascades context, reviews deliverables, makes decisions
ResearchMarket analysis, competitive teardowns, user insights, product validation
DeveloperProduction code, architecture, testing — iOS, backend, web, and generic skills
UI/UXWireframes, user flows, component specs, design tokens
QATest strategy, bug tracking, release validation
ContentIn-app copy, blog, email, store listings, help docs
MarketingGrowth strategy, campaigns, user acquisition, analytics
LegalCompliance, privacy, terms of service, IP protection (guidance, not legal advice)

All eight are peer roles bound the same way (picker or MUSTER_ROLE env var). PM is special only in what it owns (knowledge-base writes), not in how sessions bind to it.

How is this different?

CrewAI / AutoGen / LangGraphMuster
SolvesAgent communicationAgent memory and context efficiency
Built onPython libraries, API layersMarkdown files and Claude Code
RequiresCode to define agentsNo code — file-based configuration
Built fromFramework design theoryA real iOS app, mid-build
IncludesAgent orchestration primitivesOperational patterns — growth caps, cascade lag prevention, decision autonomy, pre-handoff self-review
ImprovesManual updates by maintainersProjects discover skill gaps and contribute generic skills back to the framework

Architecture

Muster uses a two-repo model.

my-project/
├── .claude/
│   ├── agents/            # Bootloaders for all 8 roles (invoke via picker or @<role>)
│   ├── skills/rebind/     # /rebind slash command (mid-session role swap)
│   ├── skills/muster/     # /muster slash command (the framework Guide)
│   ├── statusline.sh      # Status-line script (shows [muster: <role>])
│   └── settings.json      # Wires statusline + Claude Code config
├── CLAUDE.md              # Product info + project-specific rules + bootstrap routing
├── muster/                # <-- Git submodule (this repo)
├── knowledge-base/
│   ├── agent-context/     # Per-agent filtered product context (PM writes, agents read)
│   ├── product-spec.md
│   ├── orchestration-queue.md
│   ├── agent-requests.md
│   └── .muster-bind-log   # Audit trail of role binds per session
└── src/                   # Your code

Muster (this repo) contains agent roles, skills, and protocols — shared across all projects via git submodule. Your project contains product context, knowledge base, and source code. One framework, many projects.

Documentation

DocWhat it coversRead when
getting-started.mdStep-by-step setup and first sprintSetting up a new (greenfield) project
adopting-existing-project.mdReverse-discovery onboarding for projects with existing codeAdopting Muster into an existing codebase
architecture-and-design.mdArchitecture deep dive — data flow, context management, agent communicationEvaluating whether to adopt Muster
system-guide.mdTemplates, extensibility, verification checklistAdding agents, skills, or modifying the framework
operating-modes.mdThe three ways to run a sprint — Manual, Assisted, Autonomous — with cost/quality tradeoffsChoosing how to run a sprint
MIGRATING-V3-TO-V4.mdUpgrade an existing v3 project to v4 (autonomous sprint execution)Adopting v4 from an existing v3 project
MIGRATING-V2-TO-V3.mdUpgrade an existing v2 project to v3 (role-picker, status line, /rebind)Adopting v3 from an existing v2 project
MIGRATING-V1-TO-V2.mdOne-shot migration for projects set up before v2A muster/ update halted with "Pre-v2 Muster setup detected"
CHANGELOG.mdVersion history — what changed at each bumpTracing behavior to a framework version

Current framework version: 4.2 (the Guide/XO framework agent, .muster/config knobs, and autonomous-loop operability; see CHANGELOG.md). Upgrading an existing v3 project to v4? Bump the muster/ submodule and run muster/scripts/migrate-v3-to-v4.sh — see MIGRATING-V3-TO-V4.md. The 4.1 → 4.2 bump is additive: bump the submodule, then re-run migrate-v3-to-v4.sh to seed the new project-level files (the /muster front door, the test runner) — or let the Guide coach the upgrade for you (below).

Upgrading and want it done for you? From any version, bump the muster/ submodule, then tell Claude: "Read muster/MUSTER.md and act as the Guide; coach me through upgrading this project to the latest Muster." The Guide runs the migration with backups and per-step verification — see the callout atop each MIGRATING-*.md.

Stay updated

Star or watch this repo for release notifications on GitHub. For major version notes — new agents, new skills, framework improvements — subscribe to email updates. Release notes only.

License

MIT License. See LICENSE.

Ver o original no GitHub

Bom para

Relacionados