@chongyan/autospec 1.0.1 → 1.0.2

package/plugins/benchmarks/templates/pipeline/experiment-template.yaml

@@ -0,0 +1,92 @@
+# AutoSpec Pipeline Benchmark Template - Experiment
+# 适用于: 技术探索 / AI 功能验证
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "experiment-{project-name}"
+description: "实验模式 benchmark"
+
+type: pipeline
+scenario:
+  name: "技术探索实验"
+  input: "验证 {technology} 用于 {use-case} 的可行性"
+  complexity: 3
+  expectedStages:
+    - requirement
+    - design
+    - implement
+    - test
+    - deliver
+
+stageCriteria:
+  requirement:
+    expectedArtifacts:
+      - "specs/experiment/requirement.md"
+      - "specs/experiment/hypothesis.md"
+    layer1Check:
+      - "实验目标清晰"
+      - "假设可验证"
+    layer2Check:
+      - "成功指标明确"
+    maxDuration: 300
+
+  design:
+    expectedArtifacts:
+      - "specs/experiment/design.md"
+      - "specs/experiment/evaluation.md"
+    layer1Check:
+      - "实验设计合理"
+      - "评测方案可行"
+    layer2Check:
+      - "备选方案准备"
+    maxDuration: 450
+
+  implement:
+    expectedArtifacts:
+      - "src/ 原型代码"
+      - "experiments/ 实验脚本"
+    layer1Check:
+      - "原型可运行"
+      - "实验脚本可执行"
+    layer2Check:
+      - "数据收集完整"
+    maxDuration: 900
+
+  test:
+    expectedArtifacts:
+      - "experiments/results.md"
+    layer1Check:
+      - "实验执行完成"
+      - "结果记录完整"
+    layer2Check:
+      - "效果评估客观"
+    maxDuration: 600
+
+  deliver:
+    expectedArtifacts:
+      - "experiments/conclusion.md"
+    layer1Check:
+      - "结论清晰"
+    layer2Check:
+      - "知识沉淀完整"
+    maxDuration: 300
+
+pipelineMetrics:
+  expected:
+    totalDuration: 2550
+    maxRollbacks: 1
+    layer1PassRate: 85
+    layer2PassRate: 80
+    artifactCompletionRate: 90
+    successRate: 70
+    learningOutcome: 80
+
+qualityThresholds:
+  pass:
+    overallScore: 75
+    layer1PassRate: 80
+    layer2PassRate: 75
+  excellent:
+    overallScore: 85
+    layer1PassRate: 90
+    layer2PassRate: 85

package/plugins/benchmarks/templates/pipeline/hotfix-template.yaml

@@ -0,0 +1,81 @@
+# AutoSpec Pipeline Benchmark Template - Hotfix
+# 适用于: 紧急 Bug 修复模式
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "hotfix-{project-name}"
+description: "热修复模式 benchmark"
+
+type: pipeline
+scenario:
+  name: "紧急 Bug 修复"
+  input: "修复生产环境 {bug-description} 的紧急 bug"
+  complexity: 1
+  expectedStages:
+    - requirement
+    - implement
+    - test
+    - deliver
+
+stageCriteria:
+  requirement:
+    expectedArtifacts:
+      - "specs/hotfix-{id}/bug-report.md"
+    layer1Check:
+      - "bug 描述清晰"
+      - "复现步骤完整"
+    layer2Check:
+      - "影响范围明确"
+      - "快速确认"
+    maxDuration: 180
+
+  implement:
+    expectedArtifacts:
+      - "最小化修复代码"
+      - "回归测试"
+    layer1Check:
+      - "修复有效"
+      - "无新 bug 引入"
+    layer2Check:
+      - "最小化改动"
+    maxDuration: 300
+
+  test:
+    expectedArtifacts:
+      - "快速测试报告"
+    layer1Check:
+      - "bug 验证通过"
+      - "关键路径回归通过"
+    layer2Check:
+      - "测试覆盖合理"
+    maxDuration: 180
+
+  deliver:
+    expectedArtifacts:
+      - "发布说明"
+      - "复盘报告"
+    layer1Check:
+      - "发布就绪"
+    layer2Check:
+      - "复盘完整"
+    maxDuration: 120
+
+pipelineMetrics:
+  expected:
+    totalDuration: 780
+    maxRollbacks: 0
+    layer1PassRate: 95
+    layer2PassRate: 90
+    artifactCompletionRate: 100
+    fixTime: 300
+    deployTime: 120
+
+qualityThresholds:
+  pass:
+    overallScore: 90
+    layer1PassRate: 95
+    layer2PassRate: 90
+  excellent:
+    overallScore: 95
+    layer1PassRate: 100
+    layer2PassRate: 95

package/plugins/benchmarks/templates/skills/agile-iteration-template.yaml

@@ -0,0 +1,78 @@
+# AutoSpec Skill Benchmark Template - Agile-Iteration
+# 适用于: 测试 agile-iteration skill
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "skill-agile-iteration"
+description: "Agile-Iteration Skill 基准测试 - 敏捷故事循环"
+
+type: skill
+target: agile-iteration
+
+testCases:
+  - name: "single-story"
+    input:
+      context: "使用敏捷模式开发单个用户故事"
+    complexity: 1
+    expectedBehaviors:
+      - "加载故事列表"
+      - "执行故事设计"
+      - "执行故事编码"
+      - "执行故事审查"
+      - "更新故事状态"
+    expectedOutput:
+      - "stories/{id}/design.md"
+      - "代码变更"
+      - "更新后的 stories.md"
+    successCriteria:
+      - "故事设计完整"
+      - "审查通过"
+      - "状态更新正确"
+    qualityMetrics:
+      - "故事完成率 = 100%"
+      - "验证通过率 >= 90%"
+    maxDuration: 600
+
+  - name: "multi-story-sequential"
+    input:
+      context: "使用敏捷模式顺序执行多个用户故事"
+    complexity: 3
+    expectedBehaviors:
+      - "加载故事列表"
+      - "按优先级排序"
+      - "顺序执行每个故事"
+      - "集成测试"
+    expectedOutput:
+      - "各故事设计文档"
+      - "集成测试报告"
+    successCriteria:
+      - "所有故事完成"
+      - "集成测试通过"
+    qualityMetrics:
+      - "故事完成率 = 100%"
+      - "集成测试通过率 >= 90%"
+    maxDuration: 1200
+
+  - name: "parallel-stories"
+    input:
+      context: "使用敏捷模式并行执行无依赖的用户故事"
+    complexity: 5
+    expectedBehaviors:
+      - "分析故事依赖关系"
+      - "识别可并行故事"
+      - "并行执行故事"
+      - "处理代码冲突"
+    expectedOutput:
+      - "并行执行报告"
+      - "冲突检测结果"
+    successCriteria:
+      - "依赖分析正确"
+      - "并行执行成功"
+    qualityMetrics:
+      - "并行效率提升 >= 30%"
+      - "冲突检测准确率 >= 95%"
+    maxDuration: 900
+
+successCriteria:
+  passRate: 85
+  avgFieldCompletion: 90

package/plugins/benchmarks/templates/skills/benchmark-executor-template.yaml

@@ -0,0 +1,114 @@
+# AutoSpec Skill Benchmark Template - Benchmark-Executor
+# 适用于: 测试 benchmark-executor skill
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "skill-benchmark-executor"
+description: "Benchmark-Executor Skill 基准测试"
+
+type: skill
+target: benchmark-executor
+
+testCases:
+  - name: "requirement-stage-evaluation"
+    input:
+      context: "执行 requirement 阶段 benchmarks 评测"
+    complexity: 3
+    expectedBehaviors:
+      - "加载 benchmarks 文件"
+      - "检测阶段产出物"
+      - "执行 Layer1 Schema 校验"
+      - "执行 Layer2 AI 审查"
+      - "计算评分与等级"
+    expectedOutput:
+      - "评测报告"
+      - "改进建议清单"
+    successCriteria:
+      - "benchmarks 加载完整"
+      - "Layer1+Layer2 执行完成"
+      - "评分计算正确"
+    qualityMetrics:
+      - "benchmarks 加载率 = 100%"
+      - "Layer1 执行率 = 100%"
+      - "评分准确率 >= 90%"
+    maxDuration: 600
+
+  - name: "design-stage-evaluation"
+    input:
+      context: "执行 design 阶段 benchmarks 评测"
+    complexity: 3
+    expectedBehaviors:
+      - "加载 design benchmarks"
+      - "检测产出物"
+      - "执行架构设计完整性检查"
+      - "执行设计-需求一致性检查"
+    expectedOutput:
+      - "design 评测报告"
+      - "一致性检查结果"
+    successCriteria:
+      - "产出物检测完整"
+      - "一致性检查执行"
+    qualityMetrics:
+      - "产出物检测率 = 100%"
+      - "一致性检查执行率 = 100%"
+    maxDuration: 600
+
+  - name: "implementation-stage-evaluation"
+    input:
+      context: "执行 implementation 阶段 benchmarks 评测"
+    complexity: 5
+    expectedBehaviors:
+      - "加载 implementation benchmarks"
+      - "执行 Layer1 验证"
+      - "执行代码-设计一致性检查"
+      - "执行代码质量基准检查"
+    expectedOutput:
+      - "implementation 评测报告"
+      - "代码质量报告"
+      - "覆盖率报告"
+    successCriteria:
+      - "Layer1 验证真实执行"
+      - "代码质量检查完整"
+    qualityMetrics:
+      - "Layer1 执行率 = 100%"
+      - "代码检查完整率 >= 90%"
+    maxDuration: 900
+
+  - name: "multi-system-evaluation"
+    input:
+      context: "执行多系统项目 benchmarks"
+    complexity: 5
+    expectedBehaviors:
+      - "识别所有子系统"
+      - "执行接口契约一致性检查"
+      - "执行跨系统设计一致性检查"
+    expectedOutput:
+      - "多系统评测报告"
+      - "契约一致性检查结果"
+    successCriteria:
+      - "子系统识别完整"
+      - "契约一致性检查执行"
+    qualityMetrics:
+      - "子系统识别率 = 100%"
+      - "契约检查执行率 = 100%"
+    maxDuration: 900
+
+  - name: "auto-trigger-evaluation"
+    input:
+      context: "产出物变更时自动触发 benchmarks"
+    complexity: 3
+    expectedBehaviors:
+      - "检测产出物变更"
+      - "自动触发 benchmarks 评测"
+      - "执行防抖处理"
+    successCriteria:
+      - "变更检测准确"
+      - "自动触发执行"
+    qualityMetrics:
+      - "变更检测率 = 100%"
+      - "自动触发执行率 = 100%"
+    maxDuration: 300
+
+successCriteria:
+  passRate: 85
+  avgFieldCompletion: 90

package/plugins/benchmarks/templates/skills/benchmark-generator-template.yaml

@@ -0,0 +1,52 @@
+# AutoSpec Skill Benchmark Template - Benchmark-Generator
+# 适用于: 测试 benchmark-generator skill
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "skill-benchmark-generator"
+description: "Benchmark-Generator Skill 基准测试"
+
+type: skill
+target: benchmark-generator
+
+testCases:
+  - name: "generate-from-requirement"
+    input:
+      context: "根据需求文档生成 benchmark"
+    complexity: 3
+    expectedBehaviors:
+      - "读取需求文档"
+      - "提取关键场景"
+      - "生成测试用例"
+      - "生成 benchmark YAML"
+    expectedOutput:
+      - "generated-benchmark.yaml"
+      - "测试用例清单"
+    successCriteria:
+      - "场景提取完整"
+      - "benchmark 格式正确"
+    qualityMetrics:
+      - "场景覆盖率 >= 90%"
+      - "格式正确率 = 100%"
+    maxDuration: 300
+
+  - name: "generate-from-design"
+    input:
+      context: "根据设计文档生成 benchmark"
+    complexity: 3
+    expectedBehaviors:
+      - "读取设计文档"
+      - "提取 API 接口"
+      - "生成 API 测试 benchmark"
+    expectedOutput:
+      - "api-benchmark.yaml"
+    successCriteria:
+      - "API 提取完整"
+      - "benchmark 可执行"
+    qualityMetrics:
+      - "API 覆盖率 >= 90%"
+    maxDuration: 300
+
+successCriteria:
+  passRate: 85
+  avgFieldCompletion: 90

package/plugins/benchmarks/templates/skills/delivery-stage-template.yaml

@@ -0,0 +1,130 @@
+# AutoSpec Skill Benchmark Template - Delivery-Stage
+# 适用于: 测试 delivery-stage skill
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "skill-delivery-stage"
+description: "Delivery-Stage Skill 基准测试"
+
+type: skill
+target: delivery-stage
+
+testCases:
+  - name: "simple-delivery"
+    input:
+      context: "交付 {feature-name} 功能"
+    complexity: 1
+    expectedBehaviors:
+      - "执行发布检查清单"
+      - "执行一致性检查"
+      - "生成交付文档"
+      - "执行 Layer1 最终验证"
+    expectedOutput:
+      - "release-notes.md"
+      - "deployment-guide.md"
+      - "consistency-report.md"
+      - "delivery-checklist.md"
+    successCriteria:
+      - "发布检查清单完整执行"
+      - "一致性检查通过"
+      - "交付文档完整"
+    qualityMetrics:
+      - "文档完整率 >= 90%"
+      - "交付清单执行率 = 100%"
+    maxDuration: 300
+
+  - name: "multi-system-delivery"
+    input:
+      context: "交付 {feature} 系统（backend/frontend/mobile）"
+    complexity: 5
+    expectedBehaviors:
+      - "识别所有子系统"
+      - "执行各子系统发布检查"
+      - "执行集成一致性检查"
+      - "生成多系统部署方案"
+      - "生成回滚方案"
+    expectedOutput:
+      - "release-notes.md"
+      - "deployment-plan.md"
+      - "rollback-plan.md"
+      - "integration-report.md"
+    successCriteria:
+      - "所有子系统交付完整"
+      - "部署方案可行"
+      - "回滚方案完整"
+    qualityMetrics:
+      - "子系统覆盖率 = 100%"
+      - "回滚方案完整率 >= 90%"
+    maxDuration: 600
+
+  - name: "ai-delivery"
+    input:
+      context: "交付 AI 功能（含模型）"
+    complexity: 5
+    expectedBehaviors:
+      - "执行模型效果验收"
+      - "执行模型性能验收"
+      - "生成模型卡片"
+      - "生成 A/B 测试方案"
+      - "生成监控方案"
+    expectedOutput:
+      - "model-card.md"
+      - "acceptance-report.md"
+      - "ab-test-plan.md"
+      - "monitoring-plan.md"
+    successCriteria:
+      - "效果指标达标"
+      - "模型卡片完整"
+      - "监控方案完整"
+    qualityMetrics:
+      - "效果验收执行率 = 100%"
+      - "模型卡片完整率 >= 90%"
+    maxDuration: 600
+
+  - name: "safety-audit"
+    input:
+      context: "执行生产数据库迁移"
+    complexity: 3
+    expectedBehaviors:
+      - "执行安全审计"
+      - "验证数据备份方案"
+      - "验证回滚方案"
+      - "生成人工确认清单"
+    expectedOutput:
+      - "safety-audit.md"
+      - "backup-plan.md"
+      - "rollback-plan.md"
+      - "manual-checklist.md"
+    successCriteria:
+      - "安全审计执行"
+      - "备份方案验证通过"
+      - "人工确认清单完整"
+    qualityMetrics:
+      - "安全审计执行率 = 100%"
+      - "人工确认清单完整率 = 100%"
+    maxDuration: 450
+
+  - name: "practice-log"
+    input:
+      context: "记录交付后的实践日志和进化信号"
+    complexity: 3
+    expectedBehaviors:
+      - "收集实践日志"
+      - "检测进化信号"
+      - "生成进化建议"
+      - "更新 practice-log.json"
+    expectedOutput:
+      - "practice-log.json"
+      - "evolution-signal.md"
+      - "improvement-suggestion.md"
+    successCriteria:
+      - "实践日志格式正确"
+      - "进化信号检测正确"
+    qualityMetrics:
+      - "实践日志完整率 >= 90%"
+      - "进化信号检出率 = 100%"
+    maxDuration: 300
+
+successCriteria:
+  passRate: 85
+  avgFieldCompletion: 90

package/plugins/benchmarks/templates/skills/design-stage-template.yaml

@@ -0,0 +1,131 @@
+# AutoSpec Skill Benchmark Template - Design-Stage
+# 适用于: 测试 design-stage skill
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "skill-design-stage"
+description: "Design-Stage Skill 基准测试"
+
+type: skill
+target: design-stage
+
+testCases:
+  - name: "simple-design"
+    input:
+      context: "{feature-name} 功能设计"
+    complexity: 1
+    expectedBehaviors:
+      - "执行架构设计（模块划分、接口定义）"
+      - "设计 API 接口（RESTful 规范）"
+      - "设计数据库模型（ER 图、表结构）"
+      - "定义安全机制（认证、授权、加密）"
+    expectedOutput:
+      - "design.md"
+      - "api.yaml"
+      - "db-schema.sql"
+      - "security-design.md"
+    successCriteria:
+      - "架构设计合理"
+      - "API 设计符合 RESTful 规范"
+      - "数据库设计满足范式要求"
+    qualityMetrics:
+      - "设计完整率 >= 90%"
+      - "API 规范符合率 >= 95%"
+    maxDuration: 600
+
+  - name: "multi-system-design"
+    input:
+      context: "{feature} 设计，包含 {subsystems}"
+    complexity: 5
+    expectedBehaviors:
+      - "识别所有子系统"
+      - "设计整体架构"
+      - "定义系统间契约"
+      - "设计依赖顺序"
+      - "执行团队对抗审查"
+    expectedOutput:
+      - "design.md"
+      - "backend.md"
+      - "frontend.md"
+      - "contracts/api.yaml"
+    successCriteria:
+      - "识别所有子系统"
+      - "系统间契约定义清晰"
+      - "依赖顺序正确"
+    qualityMetrics:
+      - "子系统识别率 = 100%"
+      - "契约完整率 >= 90%"
+    maxDuration: 1200
+
+  - name: "ai-design"
+    input:
+      context: "AI 功能设计，使用 {model-type}"
+    complexity: 5
+    expectedBehaviors:
+      - "识别 AI/模型组件"
+      - "设计模型选型方案"
+      - "设计效果评测方案"
+      - "设计推理服务架构"
+      - "设计降级方案"
+    expectedOutput:
+      - "design.md"
+      - "ai-model.md"
+      - "evaluation.md"
+      - "fallback.md"
+    successCriteria:
+      - "AI 组件设计清晰"
+      - "效果指标可测量"
+      - "降级方案完整"
+    qualityMetrics:
+      - "AI 组件完整率 >= 90%"
+      - "评测方案完整率 >= 90%"
+    maxDuration: 900
+
+  - name: "data-modeling-design"
+    input:
+      context: "数据仓库设计，支持 BI 分析"
+    complexity: 5
+    expectedBehaviors:
+      - "识别数据源"
+      - "设计数据模型（维度建模）"
+      - "设计 ETL 流程"
+      - "设计数据质量检查机制"
+    expectedOutput:
+      - "design.md"
+      - "data-model.md"
+      - "etl-pipeline.md"
+      - "data-quality.md"
+    successCriteria:
+      - "数据源识别完整"
+      - "ETL 流程清晰可执行"
+    qualityMetrics:
+      - "数据源完整率 >= 90%"
+      - "ETL 完整率 >= 90%"
+    maxDuration: 900
+
+  - name: "high-concurrency-design"
+    input:
+      context: "秒杀系统设计，支持 10 万 QPS"
+    complexity: 5
+    expectedBehaviors:
+      - "设计缓存策略"
+      - "设计限流降级方案"
+      - "设计异步处理"
+      - "设计数据库优化"
+      - "设计容量规划"
+    expectedOutput:
+      - "design.md"
+      - "cache-design.md"
+      - "rate-limiting.md"
+      - "scaling.md"
+    successCriteria:
+      - "缓存策略合理"
+      - "限流降级方案完整"
+    qualityMetrics:
+      - "性能考虑完整率 >= 90%"
+      - "高可用设计合理率 >= 90%"
+    maxDuration: 900
+
+successCriteria:
+  passRate: 85
+  avgFieldCompletion: 90