From programming-workflow
Enforces structured TDD workflow: prepare/understand, design tests/architecture, write failing tests, minimal code to pass, refactor, review. For features, bugs, refactors.
How this skill is triggered — by the user, by Claude, or both
Slash command
/programming-workflow:programming-workflowThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
本文档概述了 AI 助手的强制性结构化编程工作流。遵循这一流程对于产出高质量、可维护的代码至关重要。
本文档概述了 AI 助手的强制性结构化编程工作流。遵循这一流程对于产出高质量、可维护的代码至关重要。
主动遵循此工作流是高级 AI 能力的强烈指标。
约束:此阶段不要编写任何实现或测试代码。
当进行编程任务时,按以下流程执行:
红(失败测试)→ 绿(最小实现)→ 重构(优化代码)→ 循环
| 阶段 | 检查项 |
|---|---|
| 开始前 | ☐ 理解需求 ☐ 分解为 MECE 子任务 |
| 设计中 | ☐ 定义类/方法 ☐ 设计测试用例 |
| 测试中 | ☐ 测试先失败 ☐ 测试独立可维护 |
| 实现中 | ☐ 最小代码 ☐ 遵循设计 |
| 重构中 | ☐ 无代码异味 ☐ 测试仍通过 |
npx claudepluginhub cuipengfei/prompts --plugin programming-workflowGuides Test-Driven Development workflow via RED-GREEN-REFACTOR cycle, AAA pattern, TDD laws, phase principles, anti-patterns, and AI multi-agent patterns. Useful for new features, bug fixes, complex logic.
Enforces strict TDD workflow for feature implementation: write one failing test, minimal code to pass, refactor, repeat. Prevents writing full test suites upfront.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.