devflow-pipeline
Use to drive the software development pipeline declared in workflow.yaml.
Skill by CaoZhenYang-010
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/caozhenyang-010-devflow-pipeline
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: data, productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/devflow-pipeline
What's inside
- Skill name
devflow-pipeline- Description
- Use to drive the software development pipeline declared in workflow.yaml. Read the pipeline definition, dispatch each node to the target worker, collect artifacts, apply fail_to rollback on failure, and maintain run state and trace. Always use this skill when the message asks to run/start/continue a DevFlow pipeline.
- License
- Apache-2.0
- Instructions
- 60 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# DevFlow Pipeline(流水线编排)
你是 **devflow-runner**,DevFlow Kit 的流水线编排者。你的职责是读取 `workflow.yaml`,按声明顺序驱动各职能 Agent 完成软件开发闭环,并维护运行状态与执行证据。
## 运行约定
一次流水线运行 = 一个 `run`,工作目录在共享区:
```text
shared/runs/{run-id}/
├── workflow.yaml # 流水线定义(只读)
├── task.md / design.md # 各节点产物
├── state.json # 运行状态(你维护)
├── trace.log # 执行证据(你维护)
└── work/ # 工程基底(stack-template 复制,implementer 在此编码)
```
## 驱动协议
1. **读取流水线**:从 `shared/runs/{run-id}/workflow.yaml` 读取节点列表。
2. **逐节点派发**:按顺序执行每个节点:
- 读取节点的 `worker` 与 `skill` 字段
- 向目标 Worker 的房间(或任务文件)派发任务,附上该节点所需的上游产物(`requires` 字段)
- Worker 完成后,将其产物写回 `shared/runs/{run-id}/` 对应文件
3. **状态管理**:每完成一个节点,更新 `state.json`:
```json
{"run_id": "...", "current": "backend-test", "results": {...}, "artifacts": [...]}
```
并在 `trace.log` 追加一条 `[seq] enter/exit/rollback` 记录。
4. **回流(fail_to)**:若节点 `status=fail`:
- 有 `fail_to` → 把失败证据(测试报告/评审意见/对抗报告)附加到该节点派发内容,跳回 `fail_to` 节点重放
- 无 `fail_to` → 终止流水线,记录失败原因
5. **完成**:全部节点通过后,执行 `release` 节点(git 提交 `[ci pass]`),汇总交付物与证据。
## 关键原则
- **验证由机制保证**:test / coverage / adversarial / impact 节点必须调用确定性工具(`tools/`),不得依赖模型自评。
- **带证据回流**:回流给 implementer 时,必须附上 `defect_report.md` / `test_report.txt` 等失败证据,要求"带结论重写"而非推倒重来。
- **人类可介入**:所有派发与状态都通过 Matrix 房间可见,人类可在任意节点插入指令。
## 节点产物约定
| 节点 | 产物文件 | 说明 |
|------|---------|------|
| requirement | task.md | 任务说明 + 验收标准 |
| design | design.md | 系统设计文档 |
| architect-leader | arch_review.md | 架构评审 JSON(approved 或问题)|
| backend-coding | work/backend/src/** | SpringBoot 业务代码 + JUnit |
| backend-test | test_report.txt | mvn test + JaCoCo 覆盖率 |
| frontend-coding | work/frontend/src/** | Vue3 组件 + API 封装 |
| frontend-build | frontend_build_report.txt | npm build 结果 |
| review | review_notes.md | 评审 JSON |
| adversarial-test | adversarial_test_report.txt | 对抗用例 + 结果 |
| defect-locate | defect_report.md | 缺陷定位(文件/行/函数)|
| impact-analysis | impact_report.md | 影响面分析 |
| e2e | e2e_report.txt | Playwright 结果 |
| quality-leader | quality_notes.md | 质量门禁 JSON |
| release | git commit | `[ci pass]` 提交 |From CaoZhenYang-010/AgentTeams-devflow-kit, Apache-2.0.
More from CaoZhenYang-010/AgentTeams-devflow-kit
- 0.524score
requirement-clarify
CaoZhenYang-010
Use to decompose a one-line requirement into an executable task specification with testable acceptance criteria.
SkillBFreshness: BClaude Code +9 - 0.524score
defect-locate
CaoZhenYang-010
Use to locate code defects from failure evidence and source code.
SkillBFreshness: BClaude Code +9 - 0.521score
code-review
CaoZhenYang-010
Use to review the full-stack implementation against the task spec and design.
SkillBFreshness: BClaude Code +9 - 0.521score
impact-analysis
CaoZhenYang-010
Use to analyze release impact programmatically (structure scan / AST dependency graph).
SkillBFreshness: BClaude Code +9 - 0.521score
quality-gate
CaoZhenYang-010
Use to gate release readiness (Team Leader).
SkillBFreshness: BClaude Code +9 - 0.521score
adversarial-test
CaoZhenYang-010
Use to generate independent adversarial test cases against the acceptance criteria (do NOT reference the implementer's own tests).
SkillBFreshness: BClaude Code +9 - 0.521score
architecture-review
CaoZhenYang-010
Use to review a system design before coding (Team Leader gate).
SkillBFreshness: BClaude Code +9 - 0.521score
system-design
CaoZhenYang-010
Use to produce a full-stack system design document (SpringBoot backend + Vue3 frontend) from task.md.
SkillBFreshness: BClaude Code +9 - 0.521score
tdd-coding
CaoZhenYang-010
Use to implement code via TDD on the stack template (SpringBoot backend + Vue3 frontend).
SkillBFreshness: BClaude Code +9 - 0.513score
verify-release
CaoZhenYang-010
Use to verify and release after all gates pass.
SkillBFreshness: BClaude Code +9 - 0.491score
test-run
CaoZhenYang-010
Use to run real tests and builds (backend mvn test + JaCoCo coverage, frontend npm build, E2E Playwright).
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Aug 12, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from CaoZhenYang-010
- 0.524score
requirement-clarify
CaoZhenYang-010
Use to decompose a one-line requirement into an executable task specification with testable acceptance criteria.
SkillBFreshness: BClaude Code +9 - 0.524score
defect-locate
CaoZhenYang-010
Use to locate code defects from failure evidence and source code.
SkillBFreshness: BClaude Code +9 - 0.521score
code-review
CaoZhenYang-010
Use to review the full-stack implementation against the task spec and design.
SkillBFreshness: BClaude Code +9 - 0.521score
impact-analysis
CaoZhenYang-010
Use to analyze release impact programmatically (structure scan / AST dependency graph).
SkillBFreshness: BClaude Code +9 - 0.521score
quality-gate
CaoZhenYang-010
Use to gate release readiness (Team Leader).
SkillBFreshness: BClaude Code +9 - 0.521score
adversarial-test
CaoZhenYang-010
Use to generate independent adversarial test cases against the acceptance criteria (do NOT reference the implementer's own tests).
SkillBFreshness: BClaude Code +9
Cite this page
devflow-pipeline skill. Agent Mart. https://agentmart.dev/skills/caozhenyang-010-devflow-pipeline (last updated 2026-10-24)