@chongyan/autospec 1.0.1 → 1.0.2

package/plugins/benchmarks/templates/commands/field-evolve-template.yaml

@@ -0,0 +1,104 @@
+# AutoSpec Command Benchmark Template - Field-Evolve
+# 适用于: 测试 field-evolve 命令（实战项目自进化）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-field-evolve"
+description: "Field-Evolve 命令基准测试"
+
+type: command
+target: field-evolve
+
+testCases:
+  - name: "quick-mode"
+    input: "--mode=quick"
+    complexity: 1
+    expectedBehaviors:
+      - "读取 config.json"
+      - "执行编译检查"
+      - "执行单元测试"
+      - "执行 Lint 检查"
+    expectedOutput:
+      - "test-result.json"
+      - "quick-report.md"
+    successCriteria:
+      - "所有检查项执行"
+      - "测试结果记录完整"
+    qualityMetrics:
+      - "检查执行率 = 100%"
+    maxDuration: 300
+
+  - name: "deep-mode"
+    input: "--mode=deep"
+    complexity: 5
+    expectedBehaviors:
+      - "扫描 benchmarks"
+      - "执行 benchmark 测试场景"
+      - "计算三维度评分"
+    expectedOutput:
+      - "deep-report.md"
+      - "evaluation.json"
+    successCriteria:
+      - "benchmark 执行完整"
+      - "三维度评分计算正确"
+    qualityMetrics:
+      - "benchmark 执行率 = 100%"
+    maxDuration: 1800
+
+  - name: "full-cycle"
+    input: "--mode=full --auto-fix"
+    complexity: 5
+    expectedBehaviors:
+      - "执行深度测试"
+      - "生成改进方案"
+      - "执行自动修复"
+      - "效果验证"
+      - "技能蒸馏"
+    expectedOutput:
+      - "full-report.md"
+      - "distilled-skills/"
+    successCriteria:
+      - "完整循环执行成功"
+      - "无退化发生"
+    qualityMetrics:
+      - "修复有效率 >= 85%"
+    maxDuration: 2400
+
+  - name: "distill-mode"
+    input: "--mode=distill"
+    complexity: 3
+    expectedBehaviors:
+      - "读取进化规则"
+      - "收集 practice-log"
+      - "执行技能蒸馏"
+      - "验证进化效果"
+    expectedOutput:
+      - "distill-report.md"
+      - "distilled-skills/"
+    successCriteria:
+      - "practice-log 完整读取"
+      - "技能提炼符合规范"
+    qualityMetrics:
+      - "技能规范率 >= 90%"
+    maxDuration: 600
+
+  - name: "generate-mode"
+    input: "--mode=generate"
+    complexity: 1
+    expectedBehaviors:
+      - "检测项目类型"
+      - "生成基础 benchmark"
+      - "生成 pipeline benchmark"
+    expectedOutput:
+      - "project-generated.yaml"
+      - "generate-report.md"
+    successCriteria:
+      - "项目类型识别正确"
+      - "benchmark 结构完整"
+    qualityMetrics:
+      - "类型识别率 = 100%"
+    maxDuration: 180
+
+successCriteria:
+  passRate: 85
+  avgDuration: 600

package/plugins/benchmarks/templates/commands/project-evolve-template.yaml

@@ -0,0 +1,104 @@
+# AutoSpec Command Benchmark Template - Project-Evolve
+# 适用于: 测试 project-evolve 命令（AI Native 项目自进化）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-project-evolve"
+description: "Project-Evolve 命令基准测试"
+
+type: command
+target: project-evolve
+
+testCases:
+  - name: "init-memory"
+    input: "--init-memory"
+    complexity: 1
+    expectedBehaviors:
+      - "扫描项目技术栈"
+      - "检测项目结构"
+      - "自动发现评测点"
+      - "创建记忆目录结构"
+    expectedOutput:
+      - ".autospec/memory/index.yaml"
+      - "初始化报告"
+    successCriteria:
+      - "技术栈识别正确"
+      - "评测点发现 >= 10 个"
+    qualityMetrics:
+      - "记忆创建率 = 100%"
+    maxDuration: 300
+
+  - name: "cruise-mode"
+    input: "--cruise"
+    complexity: 3
+    expectedBehaviors:
+      - "加载 validated 记忆"
+      - "执行代码扫描"
+      - "智能优先级排序"
+      - "L1 自动修复"
+    expectedOutput:
+      - "cruise-report.md"
+      - "auto-fixes.json"
+    successCriteria:
+      - "巡航执行完整"
+      - "自动修复正确"
+    qualityMetrics:
+      - "自动修复准确率 >= 90%"
+    maxDuration: 600
+
+  - name: "deep-evolution"
+    input: "--deep"
+    complexity: 5
+    expectedBehaviors:
+      - "六层全维度扫描"
+      - "关联分析"
+      - "生成改进方案"
+      - "区分 auto-fixable/manual-review"
+    expectedOutput:
+      - "deep-report.md"
+      - "improvements.md"
+    successCriteria:
+      - "六层扫描完整"
+      - "改进方案可执行"
+    qualityMetrics:
+      - "维度覆盖率 = 100%"
+    maxDuration: 1200
+
+  - name: "multi-system-focus"
+    input: "--focus=multi"
+    complexity: 5
+    expectedBehaviors:
+      - "识别所有子系统"
+      - "检查接口契约一致性"
+      - "检测共享逻辑抽离机会"
+      - "检测依赖循环"
+    expectedOutput:
+      - "multi-system-report.md"
+      - "dependency-graph.md"
+    successCriteria:
+      - "子系统识别完整"
+      - "契约不一致问题检出"
+    qualityMetrics:
+      - "子系统识别率 = 100%"
+    maxDuration: 900
+
+  - name: "review-mode"
+    input: "--review --last=7d"
+    complexity: 3
+    expectedBehaviors:
+      - "读取历史执行记录"
+      - "统计指标变化趋势"
+      - "分析有效改进"
+    expectedOutput:
+      - "review-report.md"
+      - "metrics-trend.md"
+    successCriteria:
+      - "历史数据完整读取"
+      - "趋势分析准确"
+    qualityMetrics:
+      - "趋势分析准确率 >= 90%"
+    maxDuration: 450
+
+successCriteria:
+  passRate: 85
+  avgDuration: 600

package/plugins/benchmarks/templates/commands/propose-template.yaml

@@ -0,0 +1,88 @@
+# AutoSpec Command Benchmark Template - Propose
+# 适用于: 测试 propose 命令（方案设计）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-propose"
+description: "Propose 命令基准测试"
+
+type: command
+target: propose
+
+testCases:
+  - name: "simple-design"
+    input: "设计 {feature-name} 功能的方案"
+    complexity: 1
+    expectedArtifacts:
+      - "specs/{feature}/design.md"
+    expectedBehaviors:
+      - "读取需求文档"
+      - "设计 API 接口"
+      - "设计数据库表"
+    successCriteria:
+      - "设计文档存在"
+      - "包含 API 设计"
+      - "包含数据库设计"
+    qualityMetrics:
+      - "API 设计完整率 >= 90%"
+      - "数据库设计合理率 >= 90%"
+    maxDuration: 300
+
+  - name: "multi-system-design"
+    input: "设计 {feature} 功能，包含 {subsystems}"
+    complexity: 5
+    expectedArtifacts:
+      - "specs/{feature}/design/overview.md"
+      - "specs/{feature}/design/backend.md"
+      - "specs/{feature}/design/frontend.md"
+      - "contracts/api.yaml"
+    expectedBehaviors:
+      - "设计整体架构"
+      - "各子系统独立设计"
+      - "定义系统间接口"
+    successCriteria:
+      - "整体架构清晰"
+      - "每个子系统有独立设计"
+      - "系统间接口定义完整"
+    qualityMetrics:
+      - "架构合理率 >= 90%"
+      - "接口完整率 >= 90%"
+    maxDuration: 600
+
+  - name: "ai-design"
+    input: "设计 AI 功能方案，使用 {model-type}"
+    complexity: 5
+    expectedArtifacts:
+      - "specs/{feature}/design.md"
+      - "specs/{feature}/evaluation.md"
+    expectedBehaviors:
+      - "设计模型选型"
+      - "设计训练流程"
+      - "设计效果评估方案"
+    successCriteria:
+      - "模型选型合理"
+      - "效果评估方案完整"
+    qualityMetrics:
+      - "技术选型合理率 >= 90%"
+      - "评估方案完整率 >= 90%"
+    maxDuration: 600
+
+  - name: "security-sensitive-design"
+    input: "设计支付系统方案"
+    complexity: 5
+    expectedBehaviors:
+      - "识别安全风险"
+      - "设计安全策略"
+      - "设计审计日志"
+    successCriteria:
+      - "安全风险识别完整"
+      - "安全策略合理"
+      - "审计日志设计完整"
+    qualityMetrics:
+      - "安全考虑完整率 >= 90%"
+      - "合规性 >= 100%"
+    maxDuration: 600
+
+successCriteria:
+  passRate: 90
+  avgDuration: 450

package/plugins/benchmarks/templates/commands/review-template.yaml

@@ -0,0 +1,124 @@
+# AutoSpec Command Benchmark Template - Review
+# 适用于: 测试 review 命令（审查交付物）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-review"
+description: "Review 命令基准测试"
+
+type: command
+target: review
+
+testCases:
+  - name: "requirement-review"
+    input: "review requirement ./specs/{feature}/requirement.md"
+    complexity: 1
+    expectedOutput:
+      - "审查结论"
+      - "逐项判定表"
+      - "blocking 问题清单"
+      - "non-blocking 建议"
+    successCriteria:
+      - "审查报告结构完整"
+      - "逐项判定有证据支撑"
+    qualityMetrics:
+      - "问题检出率 >= 90%"
+      - "误报率 <= 10%"
+    maxDuration: 300
+
+  - name: "design-review"
+    input: "review design ./specs/{feature}/design.md"
+    complexity: 3
+    expectedOutput:
+      - "审查结论"
+      - "架构设计判定"
+      - "API 设计判定"
+      - "数据库设计判定"
+      - "安全考虑判定"
+    successCriteria:
+      - "各设计维度审查完整"
+      - "设计问题识别准确"
+    qualityMetrics:
+      - "设计问题检出率 >= 85%"
+      - "建议合理率 >= 90%"
+    maxDuration: 450
+
+  - name: "code-review"
+    input: "review code ./src/"
+    complexity: 3
+    expectedOutput:
+      - "审查结论"
+      - "代码质量评分"
+      - "安全问题检测"
+      - "性能问题检测"
+    successCriteria:
+      - "代码审查维度完整"
+      - "安全漏洞识别准确"
+    qualityMetrics:
+      - "代码问题检出率 >= 85%"
+      - "安全漏洞检出率 >= 90%"
+    maxDuration: 600
+
+  - name: "test-review"
+    input: "review test ./tests/"
+    complexity: 3
+    expectedOutput:
+      - "审查结论"
+      - "测试覆盖分析"
+      - "边界条件检查"
+      - "异常路径检查"
+    successCriteria:
+      - "测试覆盖分析准确"
+      - "遗漏测试识别有效"
+    qualityMetrics:
+      - "测试遗漏检出率 >= 85%"
+      - "覆盖率评估准确率 >= 90%"
+    maxDuration: 450
+
+  - name: "git-repo-review"
+    input: "review code https://github.com/{org}/{repo} --branch {branch}"
+    complexity: 5
+    expectedBehaviors:
+      - "自动克隆 git 仓库"
+      - "指定分支检出"
+      - "执行代码审查"
+      - "生成审查报告"
+    successCriteria:
+      - "git 克隆成功"
+      - "分支检出正确"
+      - "审查报告完整"
+    qualityMetrics:
+      - "仓库审查成功率 = 100%"
+      - "问题检出率 >= 80%"
+    maxDuration: 900
+
+  - name: "auto-review"
+    input: ""
+    complexity: 1
+    expectedBehaviors:
+      - "读取 state.json 确定当前阶段"
+      - "自动推断审查类型"
+      - "自动定位审查目标"
+    successCriteria:
+      - "自动推断正确"
+      - "审查目标定位准确"
+    qualityMetrics:
+      - "自动推断准确率 >= 95%"
+    maxDuration: 300
+
+  - name: "multi-system-review"
+    input: "review code ./src/ --subsystem=backend"
+    complexity: 5
+    expectedBehaviors:
+      - "识别多系统配置"
+      - "审查指定子系统"
+    successCriteria:
+      - "子系统识别正确"
+      - "审查范围准确"
+    qualityMetrics:
+      - "子系统审查准确率 = 100%"
+    maxDuration: 600
+
+successCriteria:
+  passRate: 90
+  avgDuration: 400

package/plugins/benchmarks/templates/commands/run-template.yaml

@@ -0,0 +1,127 @@
+# AutoSpec Command Benchmark Template - Run
+# 适用于: 测试 run 命令（完整流程）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-run"
+description: "Run 命令基准测试 - 启动完整流程"
+
+type: command
+target: run
+
+testCases:
+  - name: "waterfall-simple"
+    input: "实现 {feature-name} --workflow=waterfall"
+    complexity: 1
+    expectedStages:
+      - exploration
+      - design
+      - implement
+      - test
+      - deliver
+    expectedArtifacts:
+      - "specs/{feature}/requirement.md"
+      - "specs/{feature}/design.md"
+      - "src/ 源代码"
+      - "tests/ 测试代码"
+      - "release-notes.md"
+    successCriteria:
+      - "完整执行 5 个阶段"
+      - "每个阶段通过 Layer1+Layer2 验证"
+      - "产出物完整"
+    qualityMetrics:
+      - "阶段完成率 = 100%"
+      - "Layer1 通过率 >= 90%"
+      - "Layer2 通过率 >= 80%"
+    maxDuration: 3600
+
+  - name: "agile-iteration"
+    input: "实现 {feature-name} --workflow=agile"
+    complexity: 3
+    expectedStages:
+      - exploration
+      - story-iterations
+      - integration-test
+      - deliver
+    expectedArtifacts:
+      - "specs/{feature}/requirement.md"
+      - "specs/{feature}/stories.md"
+      - "src/ 源代码"
+      - "tests/ 测试代码"
+    successCriteria:
+      - "用户故事拆分合理"
+      - "集成测试通过"
+    qualityMetrics:
+      - "故事完成率 >= 90%"
+      - "集成测试通过率 = 100%"
+    maxDuration: 1800
+
+  - name: "experiment-mode"
+    input: "实现 {feature-name} --workflow=experiment"
+    complexity: 5
+    expectedStages:
+      - exploration
+      - design
+      - prototype-implement
+      - evaluation
+      - assumption-validation
+    expectedArtifacts:
+      - "specs/{feature}/assumptions.md"
+      - "specs/{feature}/evaluation-plan.md"
+      - "specs/{feature}/evaluation-report.md"
+    successCriteria:
+      - "假设定义清晰可验证"
+      - "效果评测自动执行"
+    qualityMetrics:
+      - "假设完整率 >= 90%"
+      - "效果达标率 >= 80%"
+    maxDuration: 2700
+
+  - name: "hotfix-mode"
+    input: "修复 {bug-description} --workflow=hotfix"
+    complexity: 1
+    expectedStages:
+      - diagnosis
+      - risk-assessment
+      - quick-fix
+      - quick-review
+      - quick-deploy
+    expectedArtifacts:
+      - "specs/hotfix-{id}/hotfix-issue.md"
+      - "risk-assessment.md"
+      - "变更代码"
+      - "回归测试"
+    successCriteria:
+      - "问题定位准确"
+      - "变更最小化"
+      - "回归测试通过"
+    qualityMetrics:
+      - "修复耗时 < 15 分钟"
+    maxDuration: 900
+
+  - name: "multi-system-waterfall"
+    input: "实现 {feature}，包含后端/前端/移动端 --workflow=waterfall"
+    complexity: 5
+    expectedStages:
+      - exploration
+      - design
+      - implement
+      - test
+      - deliver
+    expectedArtifacts:
+      - "specs/{feature}/design/overview.md"
+      - "specs/{feature}/design/backend.md"
+      - "specs/{feature}/design/frontend.md"
+      - "contracts/api.yaml"
+    successCriteria:
+      - "识别所有子系统"
+      - "编码顺序正确"
+      - "集成测试通过"
+    qualityMetrics:
+      - "子系统识别率 = 100%"
+      - "依赖顺序正确率 = 100%"
+    maxDuration: 5400
+
+successCriteria:
+  passRate: 85
+  avgDuration: 1800

package/plugins/benchmarks/templates/commands/test-template.yaml

@@ -0,0 +1,149 @@
+# AutoSpec Command Benchmark Template - Test
+# 适用于: 测试 test 命令（验证+修复+评测）
+# init 后复制到 .autospec/benchmarks/ 后按需修改
+
+version: "1.0"
+name: "command-test"
+description: "Test 命令基准测试 - 统一测试命令"
+
+type: command
+target: test
+
+testCases:
+  - name: "validation-only"
+    input: "/autospec:test"
+    complexity: 1
+    expectedBehaviors:
+      - "执行编译检查"
+      - "执行测试"
+      - "执行 Lint 检查"
+      - "执行类型检查"
+    expectedOutput:
+      - "验证报告"
+      - "各项验证结果"
+    successCriteria:
+      - "验证步骤完整"
+      - "结果报告清晰"
+    qualityMetrics:
+      - "验证执行率 = 100%"
+      - "报告完整率 >= 95%"
+    maxDuration: 300
+
+  - name: "validation-with-fix"
+    input: "/autospec:test --fix"
+    complexity: 3
+    expectedBehaviors:
+      - "执行 Layer1 验证"
+      - "问题分析"
+      - "修复分类"
+      - "执行修复"
+      - "效果验证"
+    expectedOutput:
+      - "验证报告"
+      - "问题分析报告"
+      - "修复记录"
+    successCriteria:
+      - "问题分类准确"
+      - "修复有效"
+      - "退化检测正确"
+    qualityMetrics:
+      - "修复有效率 >= 85%"
+      - "退化检出率 = 100%"
+    maxDuration: 600
+
+  - name: "subsystem-test"
+    input: "/autospec:test --subsystem=backend --fix"
+    complexity: 3
+    expectedBehaviors:
+      - "读取 config.json 获取子系统配置"
+      - "只验证指定子系统"
+      - "执行子系统修复"
+    successCriteria:
+      - "子系统识别正确"
+      - "验证范围准确"
+    qualityMetrics:
+      - "子系统验证准确率 = 100%"
+    maxDuration: 450
+
+  - name: "ai-evaluation"
+    input: "/autospec:test --scope=evaluation"
+    complexity: 5
+    expectedBehaviors:
+      - "检测 AI/模型组件"
+      - "加载评测数据集"
+      - "执行效果评测"
+      - "生成评测报告"
+    expectedOutput:
+      - "评测报告"
+      - "指标结果"
+      - "Badcase 分析"
+    successCriteria:
+      - "评测方案执行完整"
+      - "指标计算准确"
+    qualityMetrics:
+      - "评测完整率 = 100%"
+      - "Badcase 检出率 >= 90%"
+    maxDuration: 900
+
+  - name: "git-repo-test"
+    input: "/autospec:test https://github.com/{org}/{repo} --branch main"
+    complexity: 5
+    expectedBehaviors:
+      - "克隆 git 仓库"
+      - "检出指定分支"
+      - "执行验证"
+    successCriteria:
+      - "git 克隆成功"
+      - "验证执行完整"
+    qualityMetrics:
+      - "仓库测试成功率 >= 95%"
+    maxDuration: 600
+
+  - name: "deep-analysis"
+    input: "/autospec:test --deep --fix"
+    complexity: 5
+    expectedBehaviors:
+      - "执行完整验证"
+      - "深度问题分析"
+      - "根因识别"
+      - "全面修复"
+    expectedOutput:
+      - "深度分析报告"
+      - "根因分析"
+      - "质量评分"
+    successCriteria:
+      - "根因识别准确"
+      - "质量提升明显"
+    qualityMetrics:
+      - "根因识别率 >= 90%"
+      - "质量提升 >= 10%"
+    maxDuration: 900
+
+  - name: "unit-test-scope"
+    input: "/autospec:test --scope=unit"
+    complexity: 1
+    expectedBehaviors:
+      - "只执行单元测试"
+      - "跳过集成测试"
+    successCriteria:
+      - "测试范围准确"
+    qualityMetrics:
+      - "单元测试覆盖率 >= 95%"
+    maxDuration: 180
+
+  - name: "integration-test-scope"
+    input: "/autospec:test --scope=integration --fix"
+    complexity: 3
+    expectedBehaviors:
+      - "执行集成测试"
+      - "识别集成问题"
+      - "修复集成问题"
+    successCriteria:
+      - "集成测试执行完整"
+    qualityMetrics:
+      - "集成测试覆盖率 = 100%"
+    maxDuration: 450
+
+successCriteria:
+  passRate: 90
+  avgDuration: 450