@kkeve/suite 0.0.1 → 0.0.2

package/doc/skills/keve-unit-gen/SKILL.md

@@ -0,0 +1,203 @@
+---
+name: keve-unit-gen
+description: 从源码导出函数自动生成 Vitest 单元测试。扫描 src/utils/、src/hooks/、src/store/ 中的纯函数和 composable，分析函数签名、边界条件和异常路径，生成 describe + it 测试文件，执行验证后输出覆盖索引。
+allowed-tools: Bash(shell:*) Bash(npx:*) Bash(npm:*) Bash(node:*) Bash(git:*) Bash(pnpm:*) Bash(yarn:*)
+---
+
+# keve-unit-gen
+
+## 定位
+
+**本技能负责**：从源码导出函数自动生成单元测试（不负责组件测试、E2E 测试）。
+
+单元测试覆盖测试金字塔最底层：纯函数、hooks、store 模块。
+运行环境为 Node.js，不需要浏览器/DOM/jsdom，毫秒级执行。
+
+***
+
+## 触发条件
+
+| 触发方式 | 场景 | 说明 |
+|----------|------|------|
+| `use_skill("keve-unit-gen")` | 开发阶段测试流水线 | 由 keve-dev-test-agent 调度 |
+| `/keve-unit-gen` 指令 | 用户主动要求生成单测 | 手动触发 |
+
+***
+
+## 输入
+
+- 项目源码中的工具函数、composable、store 模块等（由 Step 1 自动扫描识别）
+- 不硬编码目录路径，根据项目实际结构分析
+
+**排除**：
+- `.vue` 组件文件（由 keve-component-gen 负责）
+- `.spec.ts` / `.test.ts` 已有测试文件
+- 类型定义文件（`*.d.ts`）
+- 配置文件（`vite.config.*` / `tsconfig.*` / `*.config.*`）
+- `index.ts` 纯导出聚合文件
+
+***
+
+## 流程
+
+### Step 1: 扫描目标文件
+
+从项目 `src/` 目录扫描，识别导出函数的目标文件：
+
+```
+扫描路径:
+  - src/utils/**/*.ts
+  - src/hooks/**/*.ts
+  - src/store/**/*.ts
+  - src/helpers/**/*.ts
+  - src/composables/**/*.ts
+
+排除:
+  - *.vue, *.spec.ts, *.test.ts
+  - *.d.ts, *.config.ts, *.config.js
+  - index.ts (纯导出聚合)
+  - types/ 或 types.ts 目录/文件
+```
+
+输出：目标文件列表 + 每个文件的导出符号列表。
+
+### Step 2: 分析每个导出符号
+
+对每个导出函数/类/composable 分析：
+
+| 分析维度 | 识别方式 | 生成用例 |
+|---------|---------|---------|
+| 函数签名 | 参数类型 + 返回类型 | 正常路径 1 test |
+| 必填参数 | 无默认值的参数 | 缺失必填参数 1 test |
+| 可选参数 | 有 `?` 或默认值 | 有/无可选参数各 1 test |
+| 边界条件 | number → 0/-1/NaN/Infinity; string → ''/超长; array → 空数组 | 每个边界 1 test |
+| null/undefined | 参数类型含 `null | undefined` | null/undefined 输入各 1 test |
+| 异常路径 | `throw` / `reject` / `try-catch` | 异常触发 1 test |
+| 异步路径 | `async` / `Promise` | 异步正常 + 异步异常 |
+
+### Step 3: 生成测试文件
+
+**生成路径**：`keve_test_spec/util/__test__/{sourceFilename}.test.ts`
+
+**生成结构**：
+
+```typescript
+// keve_test_spec/util/__test__/format.test.ts
+import { describe, it, expect, vi } from 'vitest';
+import { truncate, formatCurrency } from '@/utils/format';
+
+describe('format', () => {
+  describe('truncate', () => {
+    it('短字符串不截断', () => {
+      expect(truncate('hello', 10)).toBe('hello');
+    });
+    it('长字符串截断加...', () => {
+      expect(truncate('hello world', 5)).toBe('he...');
+    });
+    it('刚好等于 maxLen', () => {
+      expect(truncate('hello', 5)).toBe('hello');
+    });
+    it('空字符串', () => {
+      expect(truncate('', 5)).toBe('');
+    });
+  });
+
+  describe('formatCurrency', () => {
+    it('正整数', () => {
+      expect(formatCurrency(1000)).toBe('¥1,000');
+    });
+    it('小数保留两位', () => {
+      expect(formatCurrency(1000.5)).toBe('¥1,000.50');
+    });
+  });
+});
+```
+
+**命名规则**：
+- 文件名与源文件同名，后缀 `.test.ts`
+- `describe` 第一层 = 源文件模块名
+- `describe` 第二层 = 函数名
+- `it` 描述 = 中文场景描述
+
+### Step 4: 执行验证
+
+```bash
+npx vitest run keve_test_spec/util/__test__/format.test.ts --coverage
+```
+
+- 如果通过 → 继续
+- 如果失败 → AI 分析错误，修复测试（≤2 轮）
+- 超过 2 轮仍失败 → 跳过该用例，标记为 `pending`（加 `it.todo`）
+
+### Step 5: 生成覆盖索引
+
+从 Vitest 覆盖率输出 + 测试文件分析反向生成 `keve_test_spec/util/coverage.yaml`：
+
+```yaml
+generated_at: 2026-06-16T16:05:00
+total_exports: 15
+tested_exports: 13
+untested_exports: 2
+overall_coverage: 87%
+
+files:
+  - source: src/utils/format.ts
+    test: __test__/format.test.ts
+    exports:
+      - fn: truncate
+        scenarios: [短字符串不截断, 长字符串截断加..., 刚好等于maxLen, 空字符串]
+        coverage: 100%
+      - fn: formatCurrency
+        scenarios: [正整数, 小数两位]
+        coverage: 85%
+        missing: [NaN输入, 超大数值]
+
+  - source: src/utils/validator.ts
+    test: null
+    exports:
+      - fn: validateEmail
+      - fn: validatePhone
+    missing: [全部场景]
+```
+
+***
+
+## 产出
+
+| 产出 | 路径 |
+|------|------|
+| 测试脚本 | `keve_test_spec/util/__test__/*.test.ts` |
+| 覆盖索引 | `keve_test_spec/util/coverage.yaml` |
+| 覆盖率报告 | `coverage/`（Vitest 生成） |
+
+***
+
+## 硬性约束
+
+- **不生成 .vue 组件测试**（由 keve-component-gen 负责）
+- **不需要 DOM / jsdom / 浏览器**（纯 Node.js 环境运行）
+- **优先使用 `vi.mock`**，而非 `vi.spyOn`
+- **不修改源码**（只生成测试文件）
+- **测试文件放在 `keve_test_spec/util/__test__/`**，不放在源码旁
+- **每轮修复后重新执行验证**
+- **失败的用例用 `it.todo` 标记**，不删除
+
+***
+
+## Key Decisions
+
+| ID | 决策 | 说明 |
+|----|------|------|
+| UG01 | 测试文件放在 keve_test_spec/util/ | 长期资产，跨 MR 复用（D12） |
+| UG02 | 不修改源码 | 只生成测试，不修改被测代码 |
+| UG03 | 最多 2 轮修复 | 单测修复通常简单，2 轮足够 |
+| UG04 | 失败用例用 it.todo | 保留场景信息，不删除 |
+| UG05 | 覆盖索引反向汇总 | 保证和代码一致（D09） |
+
+***
+
+## References
+
+- [unit-test-template.md](references/unit-test-template.md) — 单测生成模板
+- [unit-mock-strategy.md](references/unit-mock-strategy.md) — vi.mock 使用策略
+- [boundary-conditions.md](references/boundary-conditions.md) — 边界条件识别规则

package/doc/skills/keve-unit-gen/references/boundary-conditions.md

@@ -0,0 +1,78 @@
+# 边界条件识别规则
+
+> AI 分析函数签名时识别边界条件的参考
+
+---
+
+## 按参数类型的边界条件
+
+### number
+
+| 边界值 | 示例 |
+|--------|------|
+| 0 | `calculate(0)` |
+| -1 | `calculate(-1)` |
+| NaN | `calculate(NaN)` |
+| Infinity | `calculate(Infinity)` |
+| 超大数值 | `calculate(Number.MAX_VALUE)` |
+| 小数精度 | `calculate(0.1 + 0.2)` |
+
+### string
+
+| 边界值 | 示例 |
+|--------|------|
+| 空字符串 | `parse('')` |
+| 超长字符串 | `parse('a'.repeat(10000))` |
+| 特殊字符 | `parse('<script>alert("xss")</script>')` |
+| Unicode | `parse('你好世界🎉')` |
+| null 字符 | `parse('\0')` |
+
+### Array
+
+| 边界值 | 示例 |
+|--------|------|
+| 空数组 | `filter([])` |
+| 单元素 | `filter([1])` |
+| 重复元素 | `filter([1, 1, 1])` |
+| 混合类型 | `filter([1, 'a', null])` |
+
+### Object
+
+| 边界值 | 示例 |
+|--------|------|
+| 空对象 | `merge({})` |
+| 嵌套深对象 | `merge({ a: { b: { c: 1 } } })` |
+| 含 undefined 属性 | `merge({ a: undefined })` |
+
+### null / undefined
+
+| 边界值 | 示例 |
+|--------|------|
+| null | `parse(null)` |
+| undefined | `parse(undefined)` |
+
+---
+
+## 按返回类型的边界条件
+
+| 返回类型 | 应验证 |
+|---------|--------|
+| string | 空字符串、非空字符串 |
+| number | 0、正数、负数 |
+| boolean | true、false |
+| Array | 空数组、非空数组 |
+| Promise | resolve、reject |
+| void | 函数执行无异常 |
+
+---
+
+## 特殊场景
+
+| 场景 | 识别方式 | 测试 |
+|------|---------|------|
+| 必填参数缺失 | 无默认值 | 缺失时是否抛错 |
+| 可选参数 | `param?: T` | 有/无各 1 test |
+| 函数重载 | 多个签名 | 每个重载 1 test |
+| 泛型函数 | `<T>(x: T)` | 不同类型参数各 1 test |
+| 抛出异常 | `throw new Error()` | 异常类型 + 消息 |
+| 异步拒绝 | `reject()` | `await expect(fn()).rejects.toThrow()` |

package/doc/skills/keve-unit-gen/references/unit-mock-strategy.md

@@ -0,0 +1,71 @@
+# 单元测试 Mock 策略
+
+> AI 生成单元测试时的 vi.mock 使用指南
+
+---
+
+## Mock 优先级
+
+| 优先级 | 方式 | 适用场景 |
+|--------|------|---------|
+| 1 | `vi.mock` 整个模块 | 外部依赖（API、Store、第三方库） |
+| 2 | `vi.fn` 创建假函数 | 回调函数、事件处理器 |
+| 3 | `vi.spyOn` 监视真实方法 | 需要保留部分真实行为时 |
+| 4 | 传入测试数据 | 无外部依赖的纯函数（不需要 mock） |
+
+---
+
+## 常见 Mock 场景
+
+### Mock API 模块
+
+```typescript
+vi.mock('@/api/client', () => ({
+  get: vi.fn().mockResolvedValue({ data: [] }),
+  post: vi.fn().mockResolvedValue({ success: true }),
+}));
+```
+
+### Mock Store
+
+```typescript
+vi.mock('@/store/user', () => ({
+  useUserStore: vi.fn(() => ({
+    userInfo: { name: 'test' },
+    isLoggedIn: true,
+    login: vi.fn(),
+    logout: vi.fn(),
+  })),
+}));
+```
+
+### Mock 第三方库
+
+```typescript
+vi.mock('lodash/debounce', () => ({
+  default: (fn: Function) => fn,  // debounce 立即执行
+}));
+
+vi.mock('dayjs', () => ({
+  default: (date?: string) => ({
+    format: (fmt: string) => '2026-01-01',
+  }),
+}));
+```
+
+### Mock 环境变量
+
+```typescript
+vi.stubEnv('VITE_API_URL', 'http://localhost:3000');
+```
+
+---
+
+## 禁止做法
+
+| 禁止 | 原因 |
+|------|------|
+| Mock 被测函数本身 | 失去测试意义 |
+| Mock 纯函数 | 纯函数无副作用，直接调用 |
+| `vi.spyOn` 后 `mockRestore` 不调用 | 影响其他测试 |
+| 在 describe 顶层写 `vi.mock` 外的副作用 | 执行顺序不确定 |

package/doc/skills/keve-unit-gen/references/unit-test-template.md

@@ -0,0 +1,105 @@
+# 单元测试生成模板
+
+> AI 生成 Vitest 单元测试时的参考模板
+
+---
+
+## 基本结构
+
+```typescript
+import { describe, it, expect, vi, beforeEach, afterEach } from 'vitest';
+import { functionName } from '@/utils/module';
+
+describe('module', () => {
+  describe('functionName', () => {
+    it('正常路径描述', () => {
+      // Arrange
+      const input = validInput;
+      // Act
+      const result = functionName(input);
+      // Assert
+      expect(result).toBe(expectedValue);
+    });
+
+    it('边界条件描述', () => {
+      expect(functionName(boundaryInput)).toBe(expectedValue);
+    });
+
+    it('异常路径描述', () => {
+      expect(() => functionName(invalidInput)).toThrow(ErrorType);
+    });
+  });
+});
+```
+
+---
+
+## 纯函数模板
+
+```typescript
+describe('format', () => {
+  describe('truncate', () => {
+    it('短字符串不截断', () => {
+      expect(truncate('hello', 10)).toBe('hello');
+    });
+    it('长字符串截断加省略号', () => {
+      expect(truncate('hello world', 5)).toBe('he...');
+    });
+    it('空字符串', () => {
+      expect(truncate('', 5)).toBe('');
+    });
+  });
+});
+```
+
+---
+
+## Composable Hook 模板
+
+```typescript
+import { describe, it, expect, vi } from 'vitest';
+import { useSearch } from '@/hooks/useSearch';
+
+// 需要 @vue/test-utils 的 mountWithVitest 或 @vueuse/core 的试用
+describe('useSearch', () => {
+  it('初始化状态', () => {
+    // composable 测试需要在 setup 上下文中调用
+    // 方案1: 使用 @vue/test-utils 的 mount
+    // 方案2: 使用 vitest 的 withSetup helper
+  });
+});
+```
+
+---
+
+## 异步函数模板
+
+```typescript
+describe('fetchData', () => {
+  it('成功返回数据', async () => {
+    const mockData = { id: 1, name: 'test' };
+    vi.mock('@/api/client', () => ({
+      get: vi.fn().mockResolvedValue(mockData),
+    }));
+    const result = await fetchData(1);
+    expect(result).toEqual(mockData);
+  });
+
+  it('网络错误抛出异常', async () => {
+    vi.mock('@/api/client', () => ({
+      get: vi.fn().mockRejectedValue(new Error('Network error')),
+    }));
+    await expect(fetchData(1)).rejects.toThrow('Network error');
+  });
+});
+```
+
+---
+
+## 失败用例处理
+
+```typescript
+// 修复 2 轮后仍失败 → 改为 it.todo
+it.todo('NaN输入时的行为');
+it.todo('超大数值溢出');
+```

package/package.json

@@ -1,59 +1,81 @@
 {
   "name": "@kkeve/suite",
-  "version": "0.0.1",
-  "description": "EVE agent core runtime infrastructure — LLM client, agent loop, skill loader, context management",
+  "version": "0.0.2",
+  "description": "AI-powered full-stack test agent CLI - static analysis, unit test, component test, E2E test, execute, evaluate, and report",
   "license": "MIT",
+  "bin": {
+    "keve": "./dist/bin/keve.js"
+  },
   "main": "./dist/index.js",
   "exports": {
     ".": {
       "types": "./dist/index.d.ts",
       "default": "./dist/index.js"
     },
-    "./llm": {
-      "types": "./dist/llm/index.d.ts",
-      "default": "./dist/llm/index.js"
+    "./global-setup": {
+      "types": "./dist/core/global-setup.d.ts",
+      "default": "./dist/core/global-setup.js"
+    },
+    "./keve-report": {
+      "types": "./dist/core/keve-report.d.ts",
+      "default": "./dist/core/keve-report.js"
+    },
+    "./keve-test": {
+      "types": "./dist/core/keve-test.d.ts",
+      "default": "./dist/core/keve-test.js"
     },
-    "./context": {
-      "types": "./dist/context/index.d.ts",
-      "default": "./dist/context/index.js"
+    "./page-agent": {
+      "types": "./dist/page-agent/agent.d.ts",
+      "default": "./dist/page-agent/agent.js"
     },
-    "./agent-loop": {
-      "types": "./dist/agent-loop/index.d.ts",
-      "default": "./dist/agent-loop/index.js"
+    "./keve-aspect": {
+      "types": "./dist/core/keve-aspect.d.ts",
+      "default": "./dist/core/keve-aspect.js"
     },
-    "./skill": {
-      "types": "./dist/skill/index.d.ts",
-      "default": "./dist/skill/index.js"
+    "./action-log": {
+      "types": "./dist/core/action-log-writer.d.ts",
+      "default": "./dist/core/action-log-writer.js"
     },
-    "./agent": {
-      "types": "./dist/agent/index.d.ts",
-      "default": "./dist/agent/index.js"
+    "./learned-actions": {
+      "types": "./dist/core/learned-actions.d.ts",
+      "default": "./dist/core/learned-actions.js"
     },
-    "./tools": {
-      "types": "./dist/tools/index.d.ts",
-      "default": "./dist/tools/index.js"
+    "./spec/keve-read-doc": {
+      "types": "./dist/spec/keve-read-doc.d.ts",
+      "default": "./dist/spec/keve-read-doc.js"
     }
   },
   "files": [
-    "dist"
+    "dist",
+    "doc",
+    "README.md"
   ],
   "scripts": {
-    "build": "tsc -p tsconfig.json",
+    "build": "rm -rf dist && tsc && chmod +x dist/bin/keve.js && cp src/page-agent/system_prompt.md dist/page-agent/",
+    "dev": "tsc --watch",
     "typecheck": "tsc --noEmit",
-    "prepublishOnly": "pnpm build"
+    "keve": "node dist/bin/keve.js",
+    "prepublishOnly": ""
   },
   "repository": {
     "type": "git",
     "url": "https://github.com/bubblem1029-tech/EVE.git"
   },
   "dependencies": {
-    "glob": "^13.0.6",
-    "jsonrepair": "^3.14.0",
+    "@kkeve/core": "0.0.2",
+    "@vitest/coverage-v8": "^4.1.9",
+    "@vue/test-utils": "^2.4.6",
+    "chalk": "^4.1.0",
+    "commander": "^13.1.0",
+    "jsdom": "^26.1.0",
+    "vitest": "^4.1.9",
     "yaml": "^2.9.0",
     "zod": "^4.4.3"
   },
   "devDependencies": {
+    "@playwright/test": "^1.60.0",
     "@types/node": "^25.6.0",
+    "tsx": "^4.19.0",
     "typescript": "^6.0.2"
   }
 }

package/dist/agent/factory.d.ts

@@ -1,25 +0,0 @@
-/**
- * AgentFactory — creates execution resources for AgentLoop
- *
- * - Creates LLM, ToolRegistry, RuntimeEnvironment
- * - Provides AgentLoader for agent discovery
- * - Creates sub-agent ToolRegistry with configurable allowed/blocked tools
- */
-import type { LLMConfig } from '../llm/types';
-import type { RuntimeEnvironment } from '../agent-loop/runtime';
-import { ToolRegistry } from '../agent-loop/types';
-import { LLM } from '../llm/index';
-import { AgentLoader } from './loader';
-export declare class AgentFactory {
-    private llm;
-    private runtime;
-    private toolRegistry;
-    private agentLoader;
-    constructor(llmConfig: LLMConfig, workspaceDir?: string);
-    init(agentsDir: string): Promise<void>;
-    getLLM(): LLM;
-    getToolRegistry(): ToolRegistry;
-    getRuntime(): RuntimeEnvironment;
-    getAgentLoader(): AgentLoader;
-    createSubAgentToolRegistry(allowedTools?: string[], blockedTools?: string[]): ToolRegistry;
-}

package/dist/agent/factory.js

@@ -1,99 +0,0 @@
-"use strict";
-/**
- * AgentFactory — creates execution resources for AgentLoop
- *
- * - Creates LLM, ToolRegistry, RuntimeEnvironment
- * - Provides AgentLoader for agent discovery
- * - Creates sub-agent ToolRegistry with configurable allowed/blocked tools
- */
-var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    var desc = Object.getOwnPropertyDescriptor(m, k);
-    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
-      desc = { enumerable: true, get: function() { return m[k]; } };
-    }
-    Object.defineProperty(o, k2, desc);
-}) : (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    o[k2] = m[k];
-}));
-var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
-    Object.defineProperty(o, "default", { enumerable: true, value: v });
-}) : function(o, v) {
-    o["default"] = v;
-});
-var __importStar = (this && this.__importStar) || (function () {
-    var ownKeys = function(o) {
-        ownKeys = Object.getOwnPropertyNames || function (o) {
-            var ar = [];
-            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
-            return ar;
-        };
-        return ownKeys(o);
-    };
-    return function (mod) {
-        if (mod && mod.__esModule) return mod;
-        var result = {};
-        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
-        __setModuleDefault(result, mod);
-        return result;
-    };
-})();
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.AgentFactory = void 0;
-const path = __importStar(require("node:path"));
-const types_1 = require("../agent-loop/types");
-const index_1 = require("../llm/index");
-const runtime_1 = require("../agent-loop/runtime");
-const index_2 = require("../tools/index");
-const loader_1 = require("./loader");
-class AgentFactory {
-    llm;
-    runtime;
-    toolRegistry;
-    agentLoader;
-    constructor(llmConfig, workspaceDir) {
-        this.llm = new index_1.LLM({
-            ...llmConfig,
-            baseURL: llmConfig.baseURL || process.env.LLM_BASE_URL || '',
-            model: llmConfig.model || process.env.LLM_MODEL_NAME || process.env.LLM_MODEL || '',
-            apiKey: llmConfig.apiKey || process.env.LLM_API_KEY || '',
-        });
-        this.runtime = (0, runtime_1.detectRuntimeEnvironment)(process.cwd(), workspaceDir || path.join(process.cwd(), 'workspace'));
-        this.toolRegistry = (0, index_2.createDefaultToolRegistry)(this.runtime);
-    }
-    async init(agentsDir) {
-        this.agentLoader = new loader_1.AgentLoader(agentsDir);
-        await this.agentLoader.discover();
-    }
-    getLLM() {
-        return this.llm;
-    }
-    getToolRegistry() {
-        return this.toolRegistry;
-    }
-    getRuntime() {
-        return this.runtime;
-    }
-    getAgentLoader() {
-        return this.agentLoader;
-    }
-    createSubAgentToolRegistry(allowedTools, blockedTools) {
-        const subRegistry = new types_1.ToolRegistry();
-        const toolFilter = (allowedTools && allowedTools.length > 0)
-            ? new Set(allowedTools)
-            : null;
-        const blocked = new Set(blockedTools ?? []);
-        for (const name of this.toolRegistry.getAllToolNames()) {
-            if (blocked.has(name))
-                continue;
-            if (toolFilter && !toolFilter.has(name))
-                continue;
-            const tool = this.toolRegistry.get(name);
-            if (tool)
-                subRegistry.register(tool);
-        }
-        return subRegistry;
-    }
-}
-exports.AgentFactory = AgentFactory;

package/dist/agent/index.d.ts

@@ -1,3 +0,0 @@
-export { AgentLoader } from './loader';
-export { AgentFactory } from './factory';
-export type { AgentRegistryDefinition } from './types';