Team — Orquestrador de Agentes para Features End-to-End
bostonaholic/team
Orquestra equipes de agentes Claude para entregar features completas, da ideia ao PR
Ver no GitHubPlugin Claude Code que orquestra múltiplos agentes especializados para implementar features completas autonomamente, do conceito até PR, seguindo o workflow QRSPI (Question → Research → Design → Structure → Plan → Worktree → Implement → PR). Mantém estado persistido em artifacts e coordena 13 agentes desacoplados através de um pipeline estruturado com gates de aprovação humana.
Instalação
npx convez-ai bostonaholic/team1 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
Casos de uso
- Implementar features complexas com múltiplas disciplinas (pesquisa, design, engenharia)
- Garantir qualidade através de research desacoplado de design e gates de aprovação
- Corrigir bugs bem-compreendidos de forma acelerada com /team-fix
- Manter rastreabilidade completa de decisões em artifacts persistidos
Veredito

Skill sofisticado e altamente opinionado para times que precisam de rigor metodológico em features end-to-end. A arquitetura de agentes desacoplados com gates humanos é robusta, mas exige disciplina organizacional e compreensão profunda do workflow QRSPI para extrair valor máximo.
README
A tradução para o português deste README ainda está na fila.
Team
A Claude Code plugin that orchestrates specialized agents to autonomously implement entire features end-to-end, driven by the QRSPI workflow. The orchestrator is the main Claude Code session; it persists pipeline state as artifacts in docs/plans/ and tracks live progress with TodoWrite.
📖 Documentation: team.bostonaholic.dev
Design Philosophy
Each agent does work and returns an artifact. The orchestrator dispatches the next agent based on a phase table. Agents remain decoupled — they know nothing about each other.
Pipeline (QRSPI)
QUESTION → RESEARCH → DESIGN → STRUCTURE → PLAN → WORKTREE → IMPLEMENT → PR
- Question — Decompose intent into a full task record (
task.md) and neutral research questions (questions.md). The questioner is the only agent that ever sees the user's original description. - Research (isolated) — Parallel agents (file-finder + researcher) consume only
questions.md. They never see the task. This structurally prevents opinion-bias in research findings. - Design (human gate) — Design author asks open questions interactively, then drafts a ~200-line alignment doc. Humans review here.
- Structure (human gate) — Break the design into vertical slices with verification checkpoints. Humans review the ~2-page structure here.
- Plan — Tactical implementation plan derived from the approved structure. Read by the implementer; not human-gated.
- Worktree — Orchestrator prepares an isolated git worktree.
- Implement — Test-first (test-architect writes failing tests, mechanical gate confirms) → slice execution (implementer commits each vertical slice atomically) → adversarial verification (5 parallel reviewers + typed failure-class retry loop, max 5 rounds).
- PR — Update changelog, commit, open pull request, surface the tracking item.
Usage
/team Add rate limiting middleware to all API endpoints
For well-understood bugs, skip the QRSPI ceremony:
/team-fix Users see stale cache after profile update
Or run individual phases:
/team-question Add rate limiting middleware to all API endpoints
/team-research docs/plans/<id>/
/team-design docs/plans/<id>/
/team-structure docs/plans/<id>/
/team-plan docs/plans/<id>/
/team-worktree docs/plans/<id>/
/team-implement docs/plans/<id>/
/team-pr docs/plans/<id>/
Each command after /team-question takes the artifact directory printed by
the previous step (docs/plans/<id>/) as its single argument.
Install
claude plugin add /path/to/team
Architecture
See docs/architecture.md for the full architecture, the artifact frontmatter schema, and the phase-inference rules.
Components
- 13 agents in
agents/— decoupled workers that read predecessor artifacts fromdocs/plans/and write their outputs there - 29 entry-point + methodology skills in
skills/— slash commands and shared methodologies - 4 hooks in
hooks/— safety guards anddocs/plans/-aware compaction resilience - 1 registry at
skills/team/registry.json— phase-tagged inventory of the 13 agents - State lives in
docs/plans/<id>/*.md—<id>is<TICKET>-<topic>or<YYYY-MM-DD>-<topic>. Each artifact carries YAML frontmatter (topic,date,phase; gated artifacts also carryapproved,approved_at,revision). Live in-session coordination uses TodoWrite.