full-stack-skill / skills
CLAUDE.md0 starsChanged 8 months agoMITView on GitHub
用户通过 /dev-workflow 启动全栈开发工作流。编排器按以下 8 阶段顺序执行，每步产出作为下一步输入。 Inline 阶段（1、2、4、7）均需与用户对话。所有决策点*以结构化选项呈现*（如编号 A/B/C 让用户选择），每轮不超过 3 个问题，等待用户选择后再推进，不得跳过。
In this repository10 other files it gives its agents.
- code-developmentSkill
- dev-workflowSkill
- history-managerSkill
- program-designSkill
- requirements-analysisSkill
- self-verificationSkill
- task-breakdownSkill
- technical-reviewSkill
- technology-selectionSkill
- unit-testingSkill
The file
Files
全栈工作流入口
用户通过 /dev-workflow 启动全栈开发工作流。编排器按以下 8 阶段顺序执行，每步产出作为下一步输入。
阶段与调用方式
| 顺序 | 阶段 | Skill | 调用方式 | 说明 |
| 1 | 需求分析 | requirements-analysis | inline | 与用户对话，从原型/描述中梳理功能点，模糊处追问 |
| 2 | 技术选型 | technology-selection | inline | 先问用户全栈/仅前端/仅后端，得到答复后再产出选型清单 |
| 3 | 技术评审 | technical-review | fork/Explore | 评审架构可行性、API 契约、数据模型、风险 |
| 4 | 程序设计 | program-design | inline | 与用户确认关键假设与方案取舍后再定稿 |
| 5 | 任务拆分 | task-breakdown | fork/Plan | 将设计拆为可管理的开发单元 |
| 6 | 单元测试 | unit-testing | fork | 按任务单元设计测试用例 |
| 7 | 代码开发 | code-development | inline | 与用户确认范围/方案变更，按测试驱动实现 |
| 8 | 自我验证 | self-verification | fork/Explore | 运行测试、集成验证、质量自检 |
Inline 阶段（1、2、4、7）均需与用户对话。所有决策点以结构化选项呈现（如编号 A/B/C 让用户选择），每轮不超过 3 个问题，等待用户选择后再推进，不得跳过。
状态管理
- 所有阶段产出写入
docs/{current_iteration_id}/，文件名带 phase id。 docs/history.json记录当前迭代、各阶段 id 与状态（含创建/更新时间），避免重复生成。- 无当前迭代时先新建迭代再执行。
- 自我验证未通过时回退到代码开发修复，修复后重新验证。
- 约定见 docs-convention.md。
工具 Skill
history-manager：迭代与阶段状态管理 skill，user-invocable: false。所有阶段 skill 和编排器通过调用它来读写docs/history.json，而非直接操作文件。内置跨平台 Python 3 脚本（macOS + Windows）。- 各 skill 的详细说明见各自目录下的
SKILL.md；模板在各自templates/中。 - Frontmatter 规范见 SKILL_META.md。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.