Stackpilot
9aoyang/stackpilot
Spec → Código pronto em produção com TDD e cobertura de teste 12D
Ver no GitHubOrquestrador de sprints para Claude Code que transforma especificações em código pronto para produção. Automatiza o ciclo completo de desenvolvimento com TDD, code review em 12 dimensões e testes abrangentes, despachando sub-agentes autônomos para arquitetura, desenvolvimento e QA.
Instalação
npx convez-ai 9aoyang/stackpilot1 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
- Acelerar ciclos de desenvolvimento com automação de TDD
- Garantir qualidade com revisão arquitetônica e testes multi-dimensionais
- Orquestrar workflows complexos entre agentes especializados
Veredito

Ferramenta madura e bem-estruturada que industrializa o desenvolvimento no Claude Code com metodologia sólida (TDD, architecture review, QA sistemática). Requer integração nativa com Claude Code, limitando seu uso a esse ecossistema específico.
README
A tradução para o português deste README ainda está na fila.
Stackpilot
English | 中文
Sprint orchestration for Claude Code. Write a spec, get production-ready code — with TDD, code review, and 12-dimension test coverage.
Feature request → Design → Spec → Plan → sp-architect → sp-dev → sp-qa → Delivery
Two Layers
Portable methodology skills — work in any Agent Skills-compatible product (Cursor, VS Code Copilot, Gemini CLI, Codex, JetBrains Junie, and 25+ more):
| Skill | What it does |
|---|---|
/tdd-development | TDD cycle (RED-GREEN-REFACTOR) + verify/fix loop + rationalization blockers |
/qa-12-dimensions | Two-stage code review + 12-dimension scenario test coverage |
/architecture-review | Codebase pattern analysis → decisive architecture choice → implementation blueprint |
/systematic-debugging | 4-phase root cause investigation (observe→trace→hypothesize→fix) + red flag detection |
Stackpilot orchestration — Claude Code-only dispatch:
| Skill | What it does |
|---|---|
/stackpilot | Full sprint: tidy → resume → design → spec → plan → autonomous coding → QA → ship. Dispatches sp-* subagents via Claude Code's native Agent tool. |
/stackpilot-compete | Competitive gap analysis from power-user persona |
/stackpilot-research | Deep research reports using cross-longitudinal analysis (横纵分析法) |
/stackpilot-sync | Track and sync external skills inlined into agents |
Demo
> /stackpilot
━━━━━━━━━━━━━━━━━━━━━━━━━
Stackpilot Sprint Status
━━━━━━━━━━━━━━━━━━━━━━━━━
No active sprint.
━━━━━━━━━━━━━━━━━━━━━━━━━
What feature would you like to build?
> Add user search with fuzzy matching
Node 1: Exploring codebase...
Node 2: Design proposal ready — open http://localhost:51234/sprints/.../design-options.html or approve in terminal? (Y/n)
Node 3: Writing spec → .stackpilot/specs/2026-04-05-user-search-design.md ✓
Writing criteria → .stackpilot/specs/2026-04-05-user-search-criteria.md ✓
Spec review open at http://localhost:51234/sprints/.../spec-review.html
Node 4: Writing plan → .stackpilot/plans/2026-04-05-user-search-plan.md ✓
Live dashboard: http://localhost:51234/sprints/.../dashboard.html
Plan is ready. Proceed with coding? (Y/n)
✅ TASK-001 design search API arch → dev → QA passed (1/3)
✅ TASK-002 implement endpoint dev → QA passed (2/3)
✅ TASK-003 integration tests dev → QA passed (3/3)
Sprint complete. All tests passing.
Dev server running at: http://localhost:3000
A. Merge into main B. Push and create PR C. Leave as-is D. Discard
Install
curl -fsSL https://raw.githubusercontent.com/9aoyang/stackpilot/main/install.sh | bash
Or manually:
git clone https://github.com/9aoyang/stackpilot.git ~/Documents/github/stackpilot
bash ~/Documents/github/stackpilot/scripts/restore.sh
Requires git and Claude Code. If you use skillshare, make it the single synchronization source for shared skills.
Skills auto-update: /stackpilot checks for upstream updates once per day and pulls new skills automatically.
Config
stackpilot.config.yml is auto-generated when you first run /stackpilot. It detects your test framework automatically.
# stackpilot.config.yml (auto-generated)
qa:
coverage_threshold: 80
test_command: npm test # auto-detected from project files
Auto-detection supports: Node.js, Python, Go, Rust, Ruby, Java/Kotlin, Elixir, PHP, .NET.
Architecture
See docs/architecture.md for the full system design.
Key design decisions:
- Claude Code-native orchestration — Agent tool with
isolation: "worktree"for parallel development - Agent Skills standard — portable methodology skills work across 30+ agent products
- Progressive disclosure — SKILL.md stays lean (<500 lines), heavy content in
references/ - Plan as persistence — TaskCreate for runtime, plan files for cross-session recovery
Contributing | License
中文文档
English | 中文
面向 Claude Code 的 Sprint 编排层。写设计文档,交付生产级代码 — 含 TDD、代码审查和 12 维测试覆盖。
功能需求 → 设计讨论 → Spec → Plan → sp-architect → sp-dev → sp-qa → 交付
两层架构
便携式方法论 Skills — 在任何 Agent Skills 兼容产品中可用(Cursor、VS Code Copilot、Gemini CLI、Codex、JetBrains Junie 等 25+):
| Skill | 功能 |
|---|---|
/tdd-development | TDD 循环(RED-GREEN-REFACTOR)+ verify/fix 循环 + 合理化阻断 |
/qa-12-dimensions | 两阶段代码审查 + 12 维场景测试覆盖 |
/architecture-review | 代码库模式分析 → 唯一架构决策 → 实现蓝图 |
/systematic-debugging | 4 阶段根因调查(观察→追溯→假设→修复)+ 红旗检测 |
Stackpilot 编排 — Claude Code-only 调度:
| Skill | 功能 |
|---|---|
/stackpilot | 完整 sprint:tidy→resume→设计→spec→plan→自主编码→QA→上线。通过 Claude Code 原生 Agent 工具调度 sp-* subagents。 |
/stackpilot-compete | 以竞品重度用户视角做差距分析 |
/stackpilot-research | 横纵分析法深度研报(纵向发展史 + 横向竞品切面) |
/stackpilot-sync | 追踪和同步内联到 agent 的外部 skill |
演示
> /stackpilot
━━━━━━━━━━━━━━━━━━━━━━━━━
Stackpilot Sprint Status
━━━━━━━━━━━━━━━━━━━━━━━━━
无活跃 sprint
━━━━━━━━━━━━━━━━━━━━━━━━━
你想构建什么功能?
> 增加用户搜索,支持模糊匹配
Node 1: 探索代码库...
Node 2: 设计方案就绪 — 浏览器打开 http://localhost:51234/sprints/.../design-options.html 或终端确认?(Y/n)
Node 3: 写入 spec → .stackpilot/specs/2026-04-05-user-search-design.md ✓
写入 criteria → .stackpilot/specs/2026-04-05-user-search-criteria.md ✓
Spec 评审视图:http://localhost:51234/sprints/.../spec-review.html
Node 4: 写入 plan → .stackpilot/plans/2026-04-05-user-search-plan.md ✓
实时 Dashboard:http://localhost:51234/sprints/.../dashboard.html
计划就绪,开始编码?(Y/n)
✅ TASK-001 设计搜索 API 架构审查 → 开发 → QA 通过 (1/3)
✅ TASK-002 实现搜索接口 开发 → QA 通过 (2/3)
✅ TASK-003 集成测试 开发 → QA 通过 (3/3)
Sprint 完成,所有测试通过。
Dev server 运行中:http://localhost:3000
A. 合并到 main B. 推送并创建 PR C. 暂时保留 D. 丢弃
安装
curl -fsSL https://raw.githubusercontent.com/9aoyang/stackpilot/main/install.sh | bash
或手动安装:
git clone https://github.com/9aoyang/stackpilot.git ~/Documents/github/stackpilot
bash ~/Documents/github/stackpilot/scripts/restore.sh
需要 git 和 Claude Code。如果你使用 skillshare,应让 skillshare 成为共享 skills 的唯一同步源。
Skills 自动更新:/stackpilot 每天自动检查上游更新并拉取新 skills。
配置
stackpilot.config.yml 在首次运行 /stackpilot 时自动生成,会自动探测测试框架。
# stackpilot.config.yml(自动生成)
qa:
coverage_threshold: 80
test_command: npm test # 根据项目文件自动探测
自动探测支持:Node.js、Python、Go、Rust、Ruby、Java/Kotlin、Elixir、PHP、.NET。
架构文档
完整系统设计见 docs/architecture.zh.md。
核心设计:
- Claude Code 原生编排 — Agent tool +
isolation: "worktree"实现并行开发 - Agent Skills 标准 — 便携式方法论 Skills 可在 30+ agent 产品中使用
- 渐进式展开 — SKILL.md 精简(<500 行),重内容放
references/ - Plan 即持久层 — 运行时用 TaskCreate,跨会话用 plan 文件恢复