agentic-quality-engineering

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Agentic Quality Engineering

Agent式质量工程

<default_to_action> When implementing agentic QE or coordinating agents:

SPAWN appropriate agent(s) for the task using
```
Task
```
tool with agent type
CONFIGURE agent coordination (hierarchical/mesh/sequential)
EXECUTE with PACT principles: Proactive analysis, Autonomous operation, Collaborative feedback, Targeted risk focus
VALIDATE results through quality gates before deployment
LEARN from outcomes - store patterns in
```
aqe/learning/*
```
namespace

Quick Agent Selection:

Test generation needed →
```
qe-test-generator
```
Coverage gaps →
```
qe-coverage-analyzer
```
Quality decision →
```
qe-quality-gate
```
Security scan →
```
qe-security-scanner
```
Performance test →
```
qe-performance-tester
```
Full pipeline →
```
qe-fleet-commander
```

Critical Success Factors:

Agents amplify human expertise, not replace it
Human-in-the-loop for critical decisions
Measure: bugs caught, time saved, coverage improved </default_to_action>

<default_to_action> 在实施Agent式QE或协调Agent时：

使用
```
Task
```
工具并指定Agent类型，生成（SPAWN）适合任务的Agent
配置（CONFIGURE）Agent协调方式（层级式/网状/顺序式）
遵循PACT原则执行（EXECUTE）：主动分析、自主操作、协作反馈、针对性风险聚焦
部署前通过质量门验证（VALIDATE）结果
从结果中学习（LEARN）——将模式存储在
```
aqe/learning/*
```
命名空间中

快速Agent选择：

需要生成测试 →
```
qe-test-generator
```
覆盖缺口分析 →
```
qe-coverage-analyzer
```
质量决策 →
```
qe-quality-gate
```
安全扫描 →
```
qe-security-scanner
```
性能测试 →
```
qe-performance-tester
```
完整流水线 →
```
qe-fleet-commander
```

关键成功因素：

Agent是对人类专业能力的放大，而非替代
关键决策需保留人工介入环节
衡量指标：发现的Bug数量、节省的时间、测试覆盖率提升情况 </default_to_action>

Quick Reference Card

快速参考卡片

When to Use

适用场景

Designing autonomous testing systems
Scaling QE with intelligent agents
Implementing multi-agent coordination
Building CI/CD quality pipelines

设计自主测试系统
借助智能Agent扩展QE规模
实施多Agent协调
构建CI/CD质量流水线

PACT Principles

PACT原则

Principle	Agent Behavior	Human Role
Proactive	Analyze pre-merge, predict risk	Set guardrails
Autonomous	Execute tests, fix flaky tests	Review critical
Collaborative	Multi-agent coordination	Provide context
Targeted	Risk-based prioritization	Define risk areas

原则	Agent行为	人类角色
Proactive（主动）	分析合并前内容，预测风险	设置管控边界
Autonomous（自主）	执行测试，修复不稳定测试	审核关键结果
Collaborative（协作）	多Agent协同工作	提供上下文信息
Targeted（针对性）	基于风险优先级排序	定义风险领域

19-Agent Fleet

19-Agent集群

Category	Agents	Primary Use
Core Testing (5)	test-generator, test-executor, coverage-analyzer, quality-gate, quality-analyzer	Daily testing
Performance/Security (2)	performance-tester, security-scanner	Non-functional
Strategic (3)	requirements-validator, production-intelligence, fleet-commander	Planning
Advanced (4)	regression-risk-analyzer, test-data-architect, api-contract-validator, flaky-test-hunter	Specialized
Visual/Chaos (2)	visual-tester, chaos-engineer	Edge cases
Deployment (1)	deployment-readiness	Release
Analysis (1)	code-complexity	Maintainability

类别	Agent列表	主要用途
核心测试类（5个）	test-generator, test-executor, coverage-analyzer, quality-gate, quality-analyzer	日常测试
性能/安全类（2个）	performance-tester, security-scanner	非功能性测试
战略类（3个）	requirements-validator, production-intelligence, fleet-commander	规划阶段
进阶类（4个）	regression-risk-analyzer, test-data-architect, api-contract-validator, flaky-test-hunter	专项测试
可视化/混沌测试类（2个）	visual-tester, chaos-engineer	边缘场景测试
部署类（1个）	deployment-readiness	发布阶段
分析类（1个）	code-complexity	可维护性分析

Coordination Patterns

协调模式

Hierarchical: fleet-commander → [generators] → [executors] → quality-gate
Mesh: test-gen ↔ coverage ↔ quality (peer decisions)
Sequential: risk-analyzer → test-gen → executor → coverage → gate

Hierarchical: fleet-commander → [generators] → [executors] → quality-gate
Mesh: test-gen ↔ coverage ↔ quality (peer decisions)
Sequential: risk-analyzer → test-gen → executor → coverage → gate

Success Criteria

成功标准

✅ 10x deployment frequency with same/better quality ✅ Coverage gaps detected in real-time ✅ Bugs caught pre-production ❌ Agents acting without human oversight on critical decisions ❌ Deploying all 19 agents at once (start with 1-2)

✅ 部署频率提升10倍，且质量保持不变或更优 ✅ 实时检测到测试覆盖缺口 ✅ 生产前发现Bug ❌ Agent在无人工监督的情况下处理关键决策 ❌ 一次性部署全部19个Agent（建议从1-2个开始）

Core Concepts

核心概念

QE Evolution

QE演进历程

Stage	Approach	Limitation
Traditional	Manual everything	Human bottleneck
Automation	Scripts + fixed scenarios	Needs orchestration
Agentic	AI agents + human judgment	Requires trust-building

Core Premise: Agents amplify human expertise for 10x scale.

阶段	方式	局限性
传统阶段	全手动操作	人力瓶颈
自动化阶段	脚本+固定场景	需要编排管理
Agent驱动阶段	AI Agent+人工判断	需要建立信任机制

核心前提： Agent可将人类专业能力放大10倍，实现规模化。

Key Capabilities

核心能力

1. Intelligent Test Generation

typescript

// Agent analyzes code change, generates targeted tests
const tests = await qeTestGenerator.generate(prDiff);
// → Happy path, edge cases, error handling tests

2. Pattern Detection - Scan logs, find anomalies, correlate errors

3. Adaptive Strategy - Adjust test focus based on risk signals

4. Root Cause Analysis - Link failures to code changes, suggest fixes

1. 智能测试生成

typescript

// Agent分析代码变更，生成针对性测试
const tests = await qeTestGenerator.generate(prDiff);
// → 正常流程、边缘场景、错误处理测试

2. 模式检测 - 扫描日志，发现异常，关联错误

3. 自适应策略 - 根据风险信号调整测试重点

4. 根因分析 - 将故障与代码变更关联，建议修复方案

Agent Coordination

Agent协调

Memory Namespaces

内存命名空间

aqe/test-plan/*     - Test planning decisions
aqe/coverage/*      - Coverage analysis results
aqe/quality/*       - Quality metrics and gates
aqe/learning/*      - Patterns and Q-values
aqe/coordination/*  - Cross-agent state

aqe/test-plan/*     - 测试规划决策
aqe/coverage/*      - 覆盖分析结果
aqe/quality/*       - 质量指标与质量门
aqe/learning/*      - 模式与Q值
aqe/coordination/*  - 跨Agent状态

Memory Operations (MCP Tools)

内存操作（MCP工具）

CRITICAL: Always use

mcp__agentic-qe__memory_store

with

persist: true

for learnings.

1. Store data to persistent memory:

javascript

// Store test plan decisions (persisted to .agentic-qe/memory.db)
mcp__agentic-qe__memory_store({
  key: "aqe/test-plan/pr-123",
  namespace: "aqe/test-plan",
  value: {
    prNumber: 123,
    riskLevel: "medium",
    requiredCoverage: 85,
    testTypes: ["unit", "integration"],
    estimatedTime: 1800
  },
  persist: true,  // ⚠️ REQUIRED for cross-session persistence
  ttl: 604800     // 7 days (0 = permanent)
})

2. Retrieve prior learnings before task:

javascript

// Query patterns before starting test generation
const priorData = await mcp__agentic-qe__memory_retrieve({
  key: "aqe/learning/patterns/test-generation/*",
  namespace: "aqe/learning",
  includeMetadata: true
})

// Use patterns to guide current task
if (priorData.success) {
  console.log(`Loaded ${priorData.patterns.length} prior patterns`);
}

3. Store coverage analysis results:

javascript

mcp__agentic-qe__memory_store({
  key: "aqe/coverage/auth-module",
  namespace: "aqe/coverage",
  value: {
    moduleId: "auth-module",
    currentCoverage: 78,
    gaps: ["error-handling", "edge-cases"],
    suggestedTests: 12,
    priority: "high"
  },
  persist: true,
  ttl: 1209600  // 14 days
})

重要提示：对于学习到的模式，务必使用

mcp__agentic-qe__memory_store

并设置

persist: true

。

1. 将数据存储到持久化内存：

javascript

// 存储测试规划决策（持久化到.agentic-qe/memory.db）
mcp__agentic-qe__memory_store({
  key: "aqe/test-plan/pr-123",
  namespace: "aqe/test-plan",
  value: {
    prNumber: 123,
    riskLevel: "medium",
    requiredCoverage: 85,
    testTypes: ["unit", "integration"],
    estimatedTime: 1800
  },
  persist: true,  // ⚠️ 跨会话持久化必填
  ttl: 604800     // 7天（0表示永久）
})

2. 任务开始前获取历史学习数据：

javascript

// 生成测试前查询历史模式
const priorData = await mcp__agentic-qe__memory_retrieve({
  key: "aqe/learning/patterns/test-generation/*",
  namespace: "aqe/learning",
  includeMetadata: true
})

// 使用历史模式指导当前任务
if (priorData.success) {
  console.log(`已加载${priorData.patterns.length}个历史模式`);
}

3. 存储覆盖分析结果：

javascript

mcp__agentic-qe__memory_store({
  key: "aqe/coverage/auth-module",
  namespace: "aqe/coverage",
  value: {
    moduleId: "auth-module",
    currentCoverage: 78,
    gaps: ["error-handling", "edge-cases"],
    suggestedTests: 12,
    priority: "high"
  },
  persist: true,
  ttl: 1209600  // 14天
})

Three-Phase Memory Protocol

三阶段内存协议

For coordinated multi-agent tasks, use the STATUS → PROGRESS → COMPLETE pattern:

javascript

// PHASE 1: STATUS - Task starting
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/status",
  namespace: "aqe/coordination",
  value: { status: "running", agent: "qe-test-generator", startTime: Date.now() },
  persist: true
})

// PHASE 2: PROGRESS - Intermediate updates
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/progress",
  namespace: "aqe/coordination",
  value: { progress: 50, action: "generating-unit-tests", testsGenerated: 25 },
  persist: true
})

// PHASE 3: COMPLETE - Task finished
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/complete",
  namespace: "aqe/coordination",
  value: {
    status: "complete",
    result: "success",
    testsGenerated: 47,
    coverageAchieved: 92.3,
    duration: 15000
  },
  persist: true
})

对于多Agent协同任务，使用STATUS → PROGRESS → COMPLETE模式：

javascript

// 阶段1：STATUS - 任务启动
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/status",
  namespace: "aqe/coordination",
  value: { status: "running", agent: "qe-test-generator", startTime: Date.now() },
  persist: true
})

// 阶段2：PROGRESS - 中间更新
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/progress",
  namespace: "aqe/coordination",
  value: { progress: 50, action: "generating-unit-tests", testsGenerated: 25 },
  persist: true
})

// 阶段3：COMPLETE - 任务完成
mcp__agentic-qe__memory_store({
  key: "aqe/coordination/task-123/complete",
  namespace: "aqe/coordination",
  value: {
    status: "complete",
    result: "success",
    testsGenerated: 47,
    coverageAchieved: 92.3,
    duration: 15000
  },
  persist: true
})

Blackboard Events

黑板事件

Event	Trigger	Subscribers
`test:generated`	New tests created	executor, coverage
`coverage:gap`	Gap detected	test-generator
`quality:decision`	Gate evaluated	fleet-commander
`security:finding`	Vulnerability found	quality-gate

事件	触发条件	订阅者
`test:generated`	生成新测试	executor, coverage
`coverage:gap`	检测到覆盖缺口	test-generator
`quality:decision`	质量门评估完成	fleet-commander
`security:finding`	发现漏洞	quality-gate

Example: PR Quality Pipeline

示例：PR质量流水线

typescript

// 1. Risk analysis
const risks = await Task("Analyze PR", prDiff, "qe-regression-risk-analyzer");

// 2. Generate tests for risks
const tests = await Task("Generate tests", risks, "qe-test-generator");

// 3. Execute + analyze
const results = await Task("Run tests", tests, "qe-test-executor");
const coverage = await Task("Check coverage", results, "qe-coverage-analyzer");

// 4. Quality decision
const decision = await Task("Evaluate", {results, coverage}, "qe-quality-gate");
// → GO/NO-GO with rationale

typescript

// 1. 风险分析
const risks = await Task("Analyze PR", prDiff, "qe-regression-risk-analyzer");

// 2. 针对风险生成测试
const tests = await Task("Generate tests", risks, "qe-test-generator");

// 3. 执行+分析
const results = await Task("Run tests", tests, "qe-test-executor");
const coverage = await Task("Check coverage", results, "qe-coverage-analyzer");

// 4. 质量决策
const decision = await Task("Evaluate", {results, coverage}, "qe-quality-gate");
// → 给出通过/不通过的决策及理由

Implementation Phases

实施阶段

Phase	Duration	Goal	Agent(s)
Experiment	Weeks 1-4	Validate one use case	1 agent
Integrate	Months 2-3	CI/CD pipeline	3-4 agents
Scale	Months 4-6	Multiple use cases	8+ agents
Evolve	Ongoing	Continuous learning	Full fleet

阶段	时长	目标	使用的Agent
实验阶段	第1-4周	验证单个用例	1个Agent
集成阶段	第2-3个月	接入CI/CD流水线	3-4个Agent
规模化阶段	第4-6个月	覆盖多个用例	8个以上Agent
演进阶段	持续进行	持续学习优化	全部Agent

Phase 1 Example

阶段1示例

bash

undefined

bash

undefined

Week 1: Deploy single agent

第1周：部署单个Agent

aqe agent spawn qe-test-generator

Weeks 2-3: Generate tests for 10 PRs

第2-3周：为10个PR生成测试

Track: bugs found, test quality, review time

跟踪指标：发现的Bug数量、测试质量、评审时间

Week 4: Measure impact

第4周：衡量影响

aqe agent metrics qe-test-generator

→ Tests: 150, Bugs: 12, Time saved: 8h

→ 生成测试：150个，发现Bug：12个，节省时间：8小时

---

---

Limitations & Strengths

局限性与优势

Agents Excel At

Agent擅长的场景

Volume: Scan thousands of logs in seconds
Patterns: Find correlations humans miss
Tireless: 24/7 testing and monitoring
Speed: Instant code change analysis

批量处理：几秒内扫描数千条日志
模式识别：发现人类难以察觉的关联
持续工作：7×24小时测试与监控
快速响应：即时分析代码变更

Agents Need Humans For

Agent需要人类参与的场景

Business context and priorities
Ethical judgment and trade-offs
Creative exploration ("what if" scenarios)
Domain expertise (healthcare, finance, legal)

业务上下文与优先级定义
伦理判断与权衡决策
创造性探索（“假设”场景）
领域专业知识（医疗、金融、法律等）

Best Practices

最佳实践

Do	Don't
Start with one agent, one use case	Deploy all 18 at once
Build feedback loops early	Deploy and forget
Human reviews agent output	Auto-merge without review
Measure bugs caught, time saved	Track vanity metrics (test count)
Build trust gradually	Give full autonomy immediately

建议	避免
从单个Agent、单个用例开始	一次性部署全部18个Agent
尽早建立反馈循环	部署后不再管
人工审核Agent输出结果	无需审核直接合并
衡量发现的Bug数量、节省的时间	关注 vanity metrics（如测试数量）
逐步建立信任	立即赋予完全自主权

Trust Progression

信任演进路径

Month 1: Agent suggests → Human decides
Month 2: Agent acts → Human reviews after
Month 3: Agent autonomous on low-risk
Month 4: Agent handles critical with oversight

第1个月：Agent提出建议 → 人类做决策
第2个月：Agent执行操作 → 人类事后审核
第3个月：Agent自主处理低风险任务
第4个月：Agent在监督下处理关键任务

Agent Coordination Hints

Agent协调提示

yaml

coordination:
  topology: hierarchical
  commander: qe-fleet-commander
  memory_namespace: aqe/coordination
  blackboard_topic: qe-fleet

preload_skills:
  - agentic-quality-engineering  # Always (this skill)
  - risk-based-testing           # For prioritization
  - quality-metrics              # For measurement

agent_assignments:
  qe-test-generator: [api-testing-patterns, tdd-london-chicago]
  qe-coverage-analyzer: [quality-metrics, risk-based-testing]
  qe-security-scanner: [security-testing, risk-based-testing]
  qe-performance-tester: [performance-testing]

yaml

coordination:
  topology: hierarchical
  commander: qe-fleet-commander
  memory_namespace: aqe/coordination
  blackboard_topic: qe-fleet

preload_skills:
  - agentic-quality-engineering  # 必须加载（本技能）
  - risk-based-testing           # 用于优先级排序
  - quality-metrics              # 用于效果衡量

agent_assignments:
  qe-test-generator: [api-testing-patterns, tdd-london-chicago]
  qe-coverage-analyzer: [quality-metrics, risk-based-testing]
  qe-security-scanner: [security-testing, risk-based-testing]
  qe-performance-tester: [performance-testing]

Related Skills

Resources

资源

Agent definitions:
```
.claude/agents/
```
CLI:
```
aqe agent --help
```
Fleet status:
```
aqe fleet status
```

Success Metric: Deploy 10x more frequently with same or better quality through intelligent agent collaboration.

Agent定义：
```
.claude/agents/
```
命令行工具：
```
aqe agent --help
```
集群状态：
```
aqe fleet status
```

成功指标： 通过智能Agent协作，实现部署频率提升10倍，且质量保持不变或更优。

agentic-quality-engineering

Original

Translation

Agentic Quality Engineering

Agent式质量工程

Quick Reference Card

快速参考卡片

When to Use

适用场景

PACT Principles

PACT原则

19-Agent Fleet

19-Agent集群

Coordination Patterns

协调模式

Success Criteria

成功标准

Core Concepts

核心概念

QE Evolution

QE演进历程

Key Capabilities

核心能力

Agent Coordination

Agent协调

Memory Namespaces

内存命名空间

Memory Operations (MCP Tools)

内存操作（MCP工具）

Three-Phase Memory Protocol

三阶段内存协议

Blackboard Events

黑板事件

Example: PR Quality Pipeline

示例：PR质量流水线

Implementation Phases

实施阶段

Phase 1 Example

阶段1示例

Week 1: Deploy single agent

第1周：部署单个Agent

Weeks 2-3: Generate tests for 10 PRs

第2-3周：为10个PR生成测试

Track: bugs found, test quality, review time

跟踪指标：发现的Bug数量、测试质量、评审时间

Week 4: Measure impact

第4周：衡量影响

→ Tests: 150, Bugs: 12, Time saved: 8h

→ 生成测试：150个，发现Bug：12个，节省时间：8小时

Limitations & Strengths

局限性与优势

Agents Excel At

Agent擅长的场景

Agents Need Humans For

Agent需要人类参与的场景

Best Practices

最佳实践

Trust Progression

信任演进路径

Agent Coordination Hints

Agent协调提示

Related Skills

相关技能

Resources

资源