@tinkcarlos/skillora 0.2.0

package/.claude/skills/fullstack-developer/scripts/requirements.txt

@@ -0,0 +1,36 @@
+# Bug-fixing scripts dependencies
+# Install: pip install -r requirements.txt
+
+# === Required (for scripts to run) ===
+# None - scripts use standard library only
+
+# === Recommended formatters ===
+ruff>=0.1.0          # Fast Python linter + formatter (replaces black, isort, flake8)
+black>=23.0.0        # Python formatter (fallback if ruff unavailable)
+
+# === Recommended linters ===
+mypy>=1.0.0          # Python static type checker
+pylint>=3.0.0        # Python linter (optional, ruff preferred)
+
+# === For JavaScript/TypeScript projects ===
+# Install via npm:
+#   npm install -g prettier    # JS/TS/CSS/HTML formatter
+#   npm install -g eslint      # JS/TS linter
+#   npm install -g typescript  # TypeScript compiler (for type checking)
+
+# === For Go projects ===
+# gofmt is included with Go installation
+
+# === For Rust projects ===
+# rustfmt: rustup component add rustfmt
+
+# === Quick install commands ===
+# Python formatters/linters:
+#   pip install ruff mypy
+#
+# Node.js formatters/linters:
+#   npm install -g prettier eslint typescript
+#
+# Check dependencies:
+#   python format_code.py --check-deps
+#   python lint_check.py --check-deps

package/.claude/skills/performance-optimization/SKILL.md

@@ -0,0 +1,250 @@
+---
+name: performance-optimization
+description: |
+  Performance optimization with profiling, bottleneck analysis, and APM.
+
+  Use when:
+  - Analyzing performance bottlenecks
+  - Profiling CPU/memory usage
+  - Optimizing database queries
+  - Reducing response times
+
+  Key Features:
+  - Profiling tools (Python/Node.js/Go)
+  - Database query optimization
+  - Memory leak detection
+  - APM integration
+allowed-tools: [read, execute, grep, glob, mcp__serena__find_symbol, mcp__serena__get_symbols_overview]
+---
+
+# Performance Optimization
+
+Measure first, optimize second.
+
+## The Iron Law
+
+```
+NO OPTIMIZATION WITHOUT PROFILING. NO CHANGE WITHOUT BENCHMARK. NO DEPLOY WITHOUT LOAD TEST.
+```
+
+---
+
+## Quick Start
+
+```
+性能问题
+    │
+    ▼
+┌─────────────────────────────────────────────────────────────┐
+│ Step 1: 定位瓶颈                                             │
+│ ─────────────────────────────────────────────────────────── │
+│ CPU 密集 → Profiler (cProfile/py-spy/pprof)                 │
+│ I/O 密集 → 异步分析 + 连接池检查                             │
+│ 内存问题 → Memory Profiler + 泄漏检测                       │
+│ 数据库慢 → Query Analyzer + EXPLAIN                         │
+└─────────────────────────────────────────────────────────────┘
+    │
+    ▼
+┌─────────────────────────────────────────────────────────────┐
+│ Step 2: 优化 + 验证                                          │
+│ ─────────────────────────────────────────────────────────── │
+│ 1. 建立基准 (Baseline)                                      │
+│ 2. 实施优化                                                  │
+│ 3. 对比测试                                                  │
+│ 4. 负载测试验证                                              │
+└─────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## 🔴 性能分析工具
+
+### Python
+
+```bash
+# CPU Profiling
+python -m cProfile -o output.prof script.py
+py-spy record -o profile.svg -- python script.py
+
+# Memory Profiling
+python -m memory_profiler script.py
+tracemalloc  # 内置
+
+# 异步分析
+aiomonitor  # asyncio 监控
+```
+
+### Node.js
+
+```bash
+# CPU Profiling
+node --prof app.js
+node --inspect app.js  # Chrome DevTools
+
+# Memory
+node --expose-gc --inspect app.js
+clinic doctor -- node app.js
+```
+
+### 通用
+
+```bash
+# 负载测试
+wrk -t12 -c400 -d30s http://localhost:8000/api
+locust -f locustfile.py
+k6 run script.js
+```
+
+---
+
+## 🔴 常见性能问题
+
+### 数据库
+
+| 问题 | 检测方法 | 解决方案 |
+|------|----------|----------|
+| N+1 查询 | ORM 日志分析 | Eager loading / JOIN |
+| 缺少索引 | EXPLAIN ANALYZE | 添加索引 |
+| 全表扫描 | 慢查询日志 | 优化 WHERE 条件 |
+| 连接池耗尽 | 连接数监控 | 增加池大小 / 优化查询 |
+
+```sql
+-- 分析查询
+EXPLAIN ANALYZE SELECT * FROM users WHERE email = 'x';
+
+-- 检查索引使用
+SELECT * FROM pg_stat_user_indexes;
+```
+
+### 内存
+
+| 问题 | 检测方法 | 解决方案 |
+|------|----------|----------|
+| 内存泄漏 | 内存增长曲线 | 检查未释放引用 |
+| 大对象 | Heap snapshot | 分页 / 流式处理 |
+| 缓存膨胀 | 缓存命中率 | LRU / TTL 策略 |
+
+### API 响应
+
+| 问题 | 检测方法 | 解决方案 |
+|------|----------|----------|
+| 响应慢 | APM 追踪 | 缓存 / 异步 |
+| 阻塞调用 | 火焰图 | 异步化 |
+| 序列化慢 | Profiler | 优化数据结构 |
+
+---
+
+## 🔴 优化模式
+
+### 缓存策略
+
+```python
+# 多级缓存
+L1: 进程内缓存 (LRU, TTL=60s)
+L2: Redis (TTL=5min)
+L3: CDN (静态资源)
+
+# 缓存模式
+Cache-Aside: 读时填充
+Write-Through: 写时更新
+Write-Behind: 异步写入
+```
+
+### 异步优化
+
+```python
+# ❌ 串行
+result1 = await fetch_user(id)
+result2 = await fetch_orders(id)
+
+# ✅ 并行
+result1, result2 = await asyncio.gather(
+    fetch_user(id),
+    fetch_orders(id)
+)
+```
+
+### 批量处理
+
+```python
+# ❌ 逐条处理
+for id in ids:
+    await db.execute("INSERT INTO t VALUES (?)", (id,))
+
+# ✅ 批量处理
+await db.executemany("INSERT INTO t VALUES (?)", [(id,) for id in ids])
+```
+
+---
+
+## 🔴 性能指标 (SLI/SLO)
+
+| 指标 | 目标 | 告警阈值 |
+|------|------|----------|
+| P50 响应时间 | < 100ms | > 200ms |
+| P99 响应时间 | < 500ms | > 1s |
+| 错误率 | < 0.1% | > 1% |
+| 吞吐量 | > 1000 RPS | < 500 RPS |
+| CPU 使用率 | < 70% | > 85% |
+| 内存使用率 | < 80% | > 90% |
+
+---
+
+## 输出契约
+
+```markdown
+## Performance Optimization Report
+
+### 基准测试
+| 指标 | 优化前 | 优化后 | 提升 |
+|------|--------|--------|------|
+| P50 | 250ms | 80ms | 68% |
+| P99 | 1.2s | 300ms | 75% |
+| RPS | 500 | 1500 | 200% |
+
+### 发现的瓶颈
+1. **N+1 查询** - `api/users.py:45`
+   - 问题: 循环内查询
+   - 解决: 使用 JOIN
+
+### 实施的优化
+1. 添加 Redis 缓存
+2. 数据库索引优化
+3. 异步化外部调用
+
+### 负载测试结果
+- 工具: k6
+- 并发: 100 用户
+- 持续: 5 分钟
+- 结果: 通过
+```
+
+---
+
+## 与 fullstack-developer 集成
+
+```
+fullstack-developer
+    │
+    ├── 性能问题?
+    │       ↓
+    │   performance-optimization
+    │       │
+    │       ├── 性能分析
+    │       ├── 瓶颈定位
+    │       ├── 优化实施
+    │       └── 返回: 性能报告
+    │
+    └── 继续开发
+```
+
+---
+
+## Reference Navigation
+
+| Situation | Read This |
+|-----------|-----------|
+| Python 性能 | `references/python-profiling.md` |
+| Node.js 性能 | `references/nodejs-profiling.md` |
+| 数据库优化 | `references/database-optimization.md` |
+| 缓存策略 | `references/caching-strategies.md` |

package/.claude/skills/product-requirements/SKILL.md

@@ -0,0 +1,357 @@
+---
+name: product-requirements
+user-invocable: true
+description: |
+  Comprehensive product requirement analysis with competitive research and prioritization.
+  📋 FORMAL MODE: For complete PRD and requirement documentation (> 1 hour tasks)
+
+  Use when:
+  - "write PRD", "analyze requirements", "define user stories"
+  - "competitive analysis", "prioritize features", "scope MVP"
+  - PRD, 需求分析, 需求文档, 产品需求, 用户故事, 竞品分析, 验收标准, 用户画像
+
+  Key Features:
+  - Phase 0: Mandatory domain knowledge acquisition
+  - Answer-First Questioning (propose solutions before asking)
+  - Competitive analysis (3+ products)
+  - Structured prioritization (MoSCoW, RICE, Kano)
+
+  ⚠️ Boundary with brainstorming:
+  - brainstorming: Quick exploration, informal design (< 1 hour)
+  - product-requirements: Formal PRD, competitive analysis, user stories (> 1 hour)
+
+  Not for: technical implementation, code review, testing.
+allowed-tools: [read, write, grep, glob, web_search, exa, context7]
+---
+
+# Product Requirement Analysis
+
+Transform vague ideas into development-ready requirements using structured thinking.
+
+## ⛔ MANDATORY PRE-FLIGHT CHECK
+
+**Before ANY action, confirm:**
+
+| # | Checkpoint | Status |
+|---|------------|--------|
+| 1 | Read this SKILL.md completely? | ☐ |
+| 2 | Identified task type? (New / Optimize) | ☐ |
+| 3 | Ready to execute Phase 0 FIRST? | ☐ |
+
+**Workflow Sequence (MUST follow):**
+```
+Phase 0 (Knowledge) → Phase 1 (Analyze) → Phase 2 (Question) → Phase 3 (Document) → Phase 4 (Plan) → Phase 5 (Execute)
+```
+
+---
+
+## 🔴 Phase 0: Domain Knowledge Acquisition (ABSOLUTE FIRST)
+
+**Whether new or optimization, MUST acquire domain knowledge first!**
+
+→ **Full Protocol**: `references/knowledge-acquisition-protocol.md`
+
+### Quick Reference
+
+| Step | Action | Validation |
+|------|--------|------------|
+| 0.1 | MCP tools: `context7` + `exa` | Source date < 1 year |
+| 0.2 | Industry standards research | 2+ independent sources |
+| 0.3 | Competitor analysis (3+ products) | Pattern extraction |
+| 0.4 | Knowledge fusion | Completeness ≥ 90% |
+
+**🔴 Iron Law**: No requirement analysis without domain expertise!
+
+---
+
+## 🔴 Phase 1: Requirement Decomposition
+
+**Analyze and decompose user's request systematically.**
+
+→ **Full Protocol**: `references/requirement-decomposition-protocol.md`
+
+### Quick Analysis (每次必做)
+
+**快速三问**：
+1. **用户要什么？** → 真实目标，非表面需求
+2. **有什么约束？** → 技术栈、时间、资源
+3. **最简方案？** → 最小改动达成目标
+
+### For Optimization Requests
+
+```
+🔴 MANDATORY: Analyze existing code FIRST!
+
+1. Read current implementation
+2. Identify current state vs desired state
+3. Map affected modules
+4. Document in requirement doc as "现状分析"
+```
+
+---
+
+## 🔴 Phase 2: Answer-First Questioning
+
+**Every question MUST include AI's recommended answer!**
+
+→ **Full Protocol**: `references/answer-first-questioning-protocol.md`
+
+### Question Format (MANDATORY)
+
+```markdown
+## 问题 X: [具体问题]
+
+**AI 推荐**: [基于领域知识的推荐答案]
+
+**推荐理由**:
+- [理由1: 行业最佳实践]
+- [理由2: 竞品分析结论]
+
+**备选方案**:
+- 方案 A: [描述] - 适用场景: [...]
+- 方案 B: [描述] - 适用场景: [...]
+
+**需要确认**: 是否采用推荐方案？如有其他考虑请说明。
+```
+
+### Question Quality Gate
+
+| Check | Requirement |
+|-------|-------------|
+| Has recommendation? | ✅ Required |
+| Based on domain knowledge? | ✅ Required |
+| Avoids yes/no answers? | ✅ Required |
+| Has decision value? | ✅ Required |
+
+**🔴 Iron Law**: Questions without recommendations = Invalid!
+
+---
+
+## 🔴 Phase 3: Document Creation
+
+**Create accurate, traceable requirement document.**
+
+→ **Full Protocol**: `references/document-accuracy-protocol.md`
+
+### Document Structure
+
+```markdown
+---
+title: [项目/功能名称]
+version: 1.0.0
+status: draft
+created: YYYY-MM-DD
+knowledge_sources: [列出知识来源]
+---
+
+# [项目名称] 需求文档
+
+## 1. 领域知识摘要
+[Phase 0 获取的关键知识点]
+
+## 2. 需求分析
+### 2.1 显性需求 (用户明确说的)
+### 2.2 隐性需求 (AI 推导的，标注推导依据)
+### 2.3 现状分析 (优化需求必填)
+
+## 3. 用户确认记录
+[Phase 2 的问答记录]
+
+## 4. 功能规格
+[详细功能定义]
+
+## 5. 验收标准
+[Gherkin 格式]
+
+## 6. 开发计划
+[Phase 4 生成的计划]
+```
+
+### Traceability Rules
+
+| Requirement Type | Source Tag | Validation |
+|------------------|------------|------------|
+| 用户明确 | `[USER]` | 用户原话引用 |
+| AI 推导 | `[INFERRED]` | 推导依据必填 |
+| 行业标准 | `[STANDARD]` | 标准文档引用 |
+| 待确认 | `[PENDING]` | 必须用户确认 |
+
+---
+
+## 🔴 Phase 4: Plan Generation
+
+**Generate development plan anchored to requirement document.**
+
+→ **Full Protocol**: `references/plan-generation-protocol.md`
+
+### Plan Structure
+
+```markdown
+## 开发计划
+
+### 计划锚点 (不可变)
+- **需求文档**: [路径]
+- **成功标准**: [从需求文档提取]
+- **计划版本**: 1.0
+
+### 任务列表
+| ID | 任务 | 依赖 | 涉及 Skill | 状态 |
+|----|------|------|------------|------|
+| T1 | [任务描述] | - | `frontend-design` | pending |
+| T2 | [任务描述] | T1 | `bug-fixing` | pending |
+
+### 检查点
+- [ ] T1 完成后: 验证是否符合需求 X
+- [ ] T2 完成后: 验证是否符合需求 Y
+```
+
+---
+
+## 🔴 Phase 5: Plan Execution with Skill Orchestration
+
+**Execute plan with context isolation.**
+
+遵循 `../shared-references/skill-orchestration-protocol.md` 协议。
+
+→ **Full Protocol**: `references/plan-execution-protocol.md`
+
+### Skill Orchestration Matrix
+
+| 触发条件 | 调用 Skill | 范围约束 | 返回条件 |
+|----------|------------|----------|----------|
+| UI 开发 | `frontend-design` | 仅计划内 UI | UI 完成 |
+| Bug 发现 | `bug-fixing` | 仅阻塞性 Bug | Bug 修复 |
+| 代码完成 | `code-review` | 本次变更 | 审查通过 |
+| 测试需求 | `test-driven-development` | 计划内测试 | 测试通过 |
+
+### Context Isolation Protocol
+
+```markdown
+## 🔄 技能委托通知
+
+**主上下文锚点**:
+- 需求文档: [路径]
+- 当前计划任务: T[X]
+- 成功标准: [...]
+
+**委托详情**:
+- 目标 Skill: `[skill-name]`
+- 范围: ✅ [允许] / ❌ [禁止]
+- 退出条件: [...]
+
+---
+⏱️ 开始委托...
+
+[委托执行]
+
+---
+## ✅ 委托完成
+
+**返回主上下文**
+- 继续任务: T[X+1]
+- 计划状态: [更新]
+```
+
+### Deviation Detection
+
+| 检测项 | 触发条件 | 处理 |
+|--------|----------|------|
+| 目标漂移 | 当前动作 ≠ 计划任务 | 停止 → 询问 |
+| 范围膨胀 | 超出计划范围 | 记录 → 询问 |
+| 上下文丢失 | 忘记原始目标 | 重读锚点 |
+
+**🔴 Iron Law**: 偏离计划必须用户批准!
+
+---
+
+## Decision Tree
+
+```
+【New Requirement】
+  Phase 0: 获取领域知识 (MANDATORY)
+      ↓
+  Phase 1: 分析拆解需求
+      ↓
+  Phase 2: 带答案提问 → 用户确认
+      ↓
+  Phase 3: 编写需求文档
+      ↓
+  Phase 4: 生成开发计划
+      ↓
+  Phase 5: 按计划执行 (Skill 编排)
+
+【Optimization Requirement】
+  Phase 0: 获取领域知识 (MANDATORY)
+      ↓
+  Phase 0.5: 分析现有代码 (MANDATORY)
+      ↓
+  Phase 1: 分析优化需求
+      ↓
+  Phase 2: 带答案提问 → 用户确认
+      ↓
+  Phase 3: 更新需求文档 (标注现状 vs 目标)
+      ↓
+  Phase 4: 生成优化计划
+      ↓
+  Phase 5: 按计划执行 (Skill 编排)
+```
+
+---
+
+## Output Contract
+
+| Phase | Deliverable | Validation |
+|-------|-------------|------------|
+| 0 | 领域知识摘要 | 来源 < 1 年，2+ 独立来源 |
+| 1 | 需求分解报告 | 显性 + 隐性需求 |
+| 2 | 问答记录 | 每问题有推荐答案 |
+| 3 | 需求文档 | 可追溯性标签完整 |
+| 4 | 开发计划 | 锚点 + 检查点完整 |
+| 5 | 执行记录 | 上下文隔离日志 |
+
+---
+
+## References Navigation
+
+| File | When to Read |
+|------|--------------|
+| 🔴 `knowledge-acquisition-protocol.md` | Phase 0 |
+| 🔴 `requirement-decomposition-protocol.md` | Phase 1 |
+| 🔴 `answer-first-questioning-protocol.md` | Phase 2 |
+| 🔴 `document-accuracy-protocol.md` | Phase 3 |
+| 🔴 `plan-generation-protocol.md` | Phase 4 |
+| 🔴 `plan-execution-protocol.md` | Phase 5 |
+| `competitive-analysis-guide.md` | 竞品分析 |
+| `prioritization-frameworks.md` | RICE/MoSCoW |
+| `workflow-templates.md` | 模板参考 |
+| `feature-spec-template.md` | 功能规格 |
+| `acceptance-criteria.md` | 验收标准 |
+
+---
+
+## Definition of Done
+
+### Pre-Flight
+- [ ] 🔴 Read this SKILL.md completely
+- [ ] 🔴 Identified task type (New / Optimize)
+- [ ] 🔴 Executed Phase 0 FIRST
+
+### Knowledge Acquisition
+- [ ] 🔴 Domain knowledge acquired (MCP tools or manual)
+- [ ] 🔴 Source date < 1 year
+- [ ] 🔴 2+ independent sources verified
+
+### Questioning Quality
+- [ ] 🔴 Every question has AI recommendation
+- [ ] 🔴 Recommendations based on domain knowledge
+- [ ] 🔴 User confirmations recorded
+
+### Document Quality
+- [ ] 🔴 All requirements have traceability tags
+- [ ] 🔴 Inferred requirements have derivation basis
+- [ ] 🔴 Optimization: 现状分析 completed
+
+### Plan Execution
+- [ ] 🔴 Plan anchored to requirement document
+- [ ] 🔴 Context isolation maintained
+- [ ] 🔴 No deviation without user approval