← Back to Skills Marketplace
garkinchu

Tester Workflow

by GarkinChu · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ Security Clean
99
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install tester-workflow
Description
Complete testing workflow from requirements analysis to test case generation and review - triggers on "完整测试流程", "testing workflow", "测试工作流", "端到端测试", "全流程测试"...
README (SKILL.md)

完整测试工作流

概述

这是一个完整的测试工作流 skill,集成了从需求分析到测试用例生成和评审的全流程。一站式安装,包含所有必需的子 skills。

包含的子 Skills

本 skill 已完整包含以下4个子 skills:

  1. analyze-requirements - 需求分析(测试视角)
  2. understand-design - 设计文档理解(测试视角)
  3. generate-test-cases - 测试用例生成
  4. review-test-cases - 测试用例评审

所有子 skills 位于 included-skills/ 目录下,可以独立使用。

完整工作流

阶段1:需求分析

使用: analyze-requirements skill

输入: 需求文档
输出: 需求分析报告(6个维度)

核心原则: 不迎合用户的"简化"要求,始终保持专业标准

6个维度:

  1. 需求概述
  2. 内容提炼
  3. 关键验收标准
  4. 潜在风险点
  5. 疑问清单
  6. 一致性验证

详细文档: included-skills/analyze-requirements/


阶段2:设计文档理解

使用: understand-design skill

输入: 设计文档
输出: 问题清单(8个维度,带优先级)

核心原则: 使用测试专家的checklist,不被时间压力影响质量

8个维度:

  1. 安全性问题
  2. 数据一致性问题
  3. 边界条件问题
  4. 接口设计问题
  5. 数据库设计问题
  6. 缓存/中间件设计问题
  7. 性能问题
  8. 可测试性问题

详细文档: included-skills/understand-design/


阶段3:测试用例生成

使用: generate-test-cases skill

输入: 需求文档 + 设计文档 + 需求分析报告
输出: 测试用例(CSV格式,30-50个用例)

核心原则: 严格遵循格式要求和覆盖策略,不接受"示例就行"的降级

5个覆盖策略:

  1. 正常业务流程
  2. 异常业务流程
  3. 边界值测试
  4. 输入校验
  5. 用户体验相关

关键要求:

  • 真实CSV文件(UTF-8 BOM编码)
  • 使用 || 双竖线分隔符
  • 至少30-50个用例
  • 覆盖5个策略

详细文档: included-skills/generate-test-cases/


阶段4:测试用例评审

使用: review-test-cases skill

输入: 测试用例 + 需求文档 + 设计文档
输出: 评审报告(6个维度,带改进建议)

核心原则: 按照6个维度进行系统性评审,不接受"快速看一下"的降级

6个维度:

  1. 完整性评审
  2. 合理性评审
  3. 覆盖度评审
  4. 规范性评审
  5. 可执行性评审
  6. 改进建议

详细文档: included-skills/review-test-cases/


使用场景

场景1:完整流程(从需求到测试用例)

用户提供需求文档 
    ↓
使用 analyze-requirements 分析需求
    ↓
用户提供设计文档
    ↓
使用 understand-design 理解设计
    ↓
使用 generate-test-cases 生成测试用例
    ↓
使用 review-test-cases 评审测试用例
    ↓
输出最终测试用例

场景2:部分流程(只需要生成测试用例)

用户提供需求文档 + 设计文档
    ↓
使用 generate-test-cases 生成测试用例
    ↓
输出测试用例

场景3:部分流程(只需要评审测试用例)

用户提供测试用例 + 需求文档
    ↓
使用 review-test-cases 评审测试用例
    ↓
输出评审报告

工作流程图

┌─────────────────┐
│   需求文档      │
└────────┬────────┘
         │
         ↓
┌─────────────────────────┐
│ analyze-requirements    │
│ (需求分析)              │
└────────┬────────────────┘
         │
         ↓
┌─────────────────┐
│ 需求分析报告    │
└────────┬────────┘
         │
         ↓
┌─────────────────┐
│   设计文档      │
└────────┬────────┘
         │
         ↓
┌─────────────────────────┐
│ understand-design       │
│ (设计文档理解)          │
└────────┬────────────────┘
         │
         ↓
┌─────────────────┐
│   问题清单      │
└────────┬────────┘
         │
         ↓ (澄清问题后)
┌─────────────────────────┐
│ generate-test-cases     │
│ (测试用例生成)          │
└────────┬────────────────┘
         │
         ↓
┌─────────────────┐
│ 测试用例(CSV)   │
└────────┬────────┘
         │
         ↓
┌─────────────────────────┐
│ review-test-cases       │
│ (测试用例评审)          │
└────────┬────────────────┘
         │
         ↓
┌─────────────────┐
│   评审报告      │
└────────┬────────┘
         │
         ↓ (修改后)
┌─────────────────┐
│ 最终测试用例    │
└─────────────────┘

核心原则

1. 不降低标准

  • 不迎合用户的"简化"、"快速"、"示例"要求
  • 始终保持专业标准
  • 系统性分析,不做表面工作

2. 系统性方法

  • analyze-requirements:6个维度
  • understand-design:8个维度
  • generate-test-cases:5个覆盖策略
  • review-test-cases:6个维度

3. 量化分析

  • 提供覆盖率百分比
  • 标注优先级(P0/P1/P2)
  • 统计用例数量
  • 评估风险等级

4. 具体可执行

  • 提供标准模板
  • 给出具体建议
  • 估算工作量
  • 明确下一步行动

快速开始

1. 完整流程示例

参见:examples/full-workflow.md

2. 各阶段详细文档

3. 工作流程指南

参见:reference/workflow-guide.md


文件结构

tester-workflow/
├── README.md                          # 本文件
├── SKILL.md                           # 主 skill 文件
├── included-skills/                   # 包含的子 skills
│   ├── analyze-requirements/          # 需求分析
│   ├── understand-design/             # 设计理解
│   ├── generate-test-cases/           # 用例生成
│   └── review-test-cases/             # 用例评审
├── examples/
│   └── full-workflow.md              # 完整工作流示例
└── reference/
    └── workflow-guide.md             # 工作流程指南

底线

专业标准不应因用户的时间压力或简化要求而降低。

每个阶段都有明确的质量标准:

  • 需求分析:6个维度全覆盖
  • 设计理解:8个维度系统性提问
  • 用例生成:30-50个用例,5个策略全覆盖
  • 用例评审:6个维度全面评审

如果你发现自己在想"快速做一下就行",立即停止。必须按照标准流程执行。

Usage Guidance
This is an instruction-only skill that bundles documentation and templates for a complete testing workflow; it does not ask for credentials or install code, which makes it internally coherent. Before installing: 1) review the raw SKILL.md (and any omitted files) for hidden characters or unexpected instructions because the scanner flagged unicode control characters; 2) verify included example/template files do not contain customer-sensitive data; 3) confirm how your platform's 'Write' tool works (what path the CSVs are written to and whether files are accessible externally); 4) if you will let the agent operate autonomously, be aware the skill enforces strict refusal language (it will insist on full workflows rather than shortcuts) — that is a design choice, not a security issue. If you want extra assurance, run the skill in a sandboxed agent session first and check generated outputs and any side-effects.
Capability Analysis
Type: OpenClaw Skill Name: tester-workflow Version: 1.0.1 The skill bundle defines a professional software testing workflow across multiple files (SKILL.md, included-skills/). It employs prescriptive instructions to force the AI agent to adhere to specific quality standards (e.g., 6-dimension analysis) even if the user requests a 'quick' or 'simplified' version. While these instructions are designed to override certain user preferences, they are strictly limited to the domain of software testing quality and do not exhibit malicious intent, data exfiltration, or unauthorized system manipulation.
Capability Tags
requires-oauth-token
Capability Assessment
Purpose & Capability
Name/description (end-to-end testing workflow) align with the contents: SKILL.md plus four included sub-skills for requirements analysis, design understanding, test-case generation, and review. No unrelated env vars, binaries, or config paths are requested.
Instruction Scope
Runtime instructions are purely procedural and document-heavy (generate reports, CSVs, checklists). They ask the agent to require user-provided documents and to produce files (mentions using a 'Write tool' to generate real CSVs). This is within scope, but the SKILL.md enforces strict, prescriptive behavior (refuse 'quick' requests) which grants the skill strong operational rules. Also a regex pre-scan flagged unicode-control-chars in SKILL.md — this could hide content; review raw files if concerned.
Install Mechanism
No install spec and no code files to execute. Instruction-only skills pose minimal installation risk because nothing will be downloaded or written by an installer step.
Credentials
Skill declares no required environment variables, no credentials, and no config paths. All required inputs are user-provided documents (requirements, design, test cases), which is proportional to the stated purpose.
Persistence & Privilege
Flags show default privileges (always: false, model invocation enabled). The skill does not request permanent agent presence or special privileges, and does not attempt to modify other skills or agent-wide config in the provided files.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install tester-workflow
  3. After installation, invoke the skill by name or use /tester-workflow
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Renamed from "testing-workflow" to "tester-workflow" throughout the documentation and file structure. - Renamed main metadata file from "skill.md" to "SKILL.md". - Updated all file paths, examples, and documentation to reflect the new naming.
v1.0.0
Initial release of the "testing-workflow" skill with end-to-end testing process. - Provides a comprehensive, modular workflow for testing: from requirements analysis and design understanding to test case generation and review. - Includes four integrated sub-skills: analyze-requirements, understand-design, generate-test-cases, and review-test-cases—each available for independent use. - Offers detailed, step-by-step documentation and templates for each workflow phase with strict professional standards. - Supports different usage scenarios: full process, test-case generation only, or test-case review only. - All instructions, principles, and process diagrams are included for clarity and quick onboarding.
Metadata
Slug tester-workflow
Version 1.0.1
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Tester Workflow?

Complete testing workflow from requirements analysis to test case generation and review - triggers on "完整测试流程", "testing workflow", "测试工作流", "端到端测试", "全流程测试"... It is an AI Agent Skill for Claude Code / OpenClaw, with 99 downloads so far.

How do I install Tester Workflow?

Run "/install tester-workflow" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Tester Workflow free?

Yes, Tester Workflow is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Tester Workflow support?

Tester Workflow is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Tester Workflow?

It is built and maintained by GarkinChu (@garkinchu); the current version is v1.0.1.

💬 Comments