agentleFS
Ron-daliRon-dali/vibe-coding-rules/pipeline-init/SKILL.md

pipeline-init

Skill4 starsChanged 4 months agoApache-2.0View on GitHub

Conversational project info collection → dependency detection → scaffold creation → pipeline template fork → ready in one go. V2.6 adds Step -1: Platform Detection → Memory Self-Heal → Project State Detection with three-path handling. V2.7 softens Memory seeding: discover + suggest, not force-write. Cross-platform compatibility first. V2.8 fixes Path C: adds user confirmation + language selection. V2.9 去模板化:Path C 从固定模板创建改为项目扫描→智能提案→用户确认,只创建项目真正需要的目录。 Activate when any of the following occurs: - User says "I want to build X project", "Help me init…

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. johnrtipton/pipeline-skills★ 4same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository6 other files it gives its agents.

The file

Show its rules and text

We split this skill into the rules it gives an agent

24 rules. About 0 are common, 11 are distinctive, and 13 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 13

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the file at memoryPath (from Step -1.0)

    line 45
  2. Distinctive:

    Check for 5 critical sections (using fuzzy markdown heading match):

    line 46
  3. Distinctive:

    If ANY section is missing → detect and report (V2.7: do NOT auto-write without confirmation)

    line 52
  4. Very specific:

    Only write to memoryPath after explicit user confirmation

    line 54
  5. Distinctive:

    If file does not exist → ask: "没有检测到 Memory 文件,要创建吗?(是 / 先不要)"

    line 55
  6. Distinctive:

    After memory self-heal, detect project state:

    line 68
  7. Very specific:

    Check if project root has code files (.js, .html, .py, .ts, etc.) → NO → Path A: New Project

    line 70
  8. Very specific:

    Check if 开发文档/ directory exists (or configured devDocs path) → NO → Path C: Existing Project Without Docs

    line 72
  9. Very specific:

    Check if 开发文档/00-必读/开发文档管理.md exists → YES → Path B: Existing Project With Docs → NO → Path C: Existing Project Without Docs

    line 74
  10. Very specific:

    Run DOC_SPEC compliance scan:

    line 84
  11. Very specific:

    Check all docs in 开发文档/ have 📋 管理规范见 reference in header

    line 85
  12. Distinctive:

    Check no orphan files in wrong directories

    line 86
  13. Distinctive:

    Check file size limits (≤500 lines)

    line 87
  14. Very specific:

    Verify 📖 references point to existing docs

    line 88
  15. Distinctive:

    Verify management doc index is up to date

    line 89
  16. Distinctive:

    Ask user: "Found some documentation issues.

    line 91
  17. Distinctive:

    If user says "fix now" → repair issues (add 📋, split large files, fix broken links)

    line 92
  18. Very specific:

    Language selection first: Run Phase 0 of references/init-dialog.md to ask user's preferred language (中文 / English / 日本語 / 其他).

    line 104
  19. Very specific:

    Follow references/init-dialog.md Phase 0 to ask the user's preferred language.

    line 190
  20. Very specific:

    Follow references/init-dialog.md template to progressively collect:

    line 194
  21. Very specific:

    Check environment: Node.js version, npm, Git, additional deps per tech stack.

    line 199
  22. Very specific:

    Create .ai-pipeline/ with pipeline.json, VERSION (2.9.0), and self-check references.

    line 204
  23. Distinctive:

    Write README.md with proper language and no hardcoded versions.

    line 206
  24. Distinctive:

    If memory was self-healed in Step -1.a, mention which sections were repaired.

    line 220

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

pipeline-init — Project Initialization Wizard V2.9

Conversational project info collection → dependency detection → scaffold creation → pipeline template fork → ready in one go. V2.6 adds Step -1: Platform Detection → Memory Self-Heal → Project State Detection with three-path handling. V2.7 softens Memory seeding: discover + suggest, not force-write. Cross-platform compatibility first. V2.8 fixes Path C: adds user confirmation + language selection. V2.9 去模板化:Path C 从固定模板创建改为项目扫描→智能提案→用户确认,只创建项目真正需要的目录。

Trigger Conditions

Activate when any of the following occurs:

  • User says "I want to build X project", "Help me init the pipeline", "New project"
  • Memory trigger hits new-project intent
  • User mentions starting a new project
  • V2.6: Skill first loaded in ANY project (runs Step -1 for memory self-heal)

Workflow

Step -1: Pre-Initialization Detection(V2.6 NEW — Run first on every skill load)

This step runs BEFORE any other workflow. It is idempotent — safe to run repeatedly.

Step -1.0: Platform Detection(确定 Memory 文件路径)

Detect which AI coding tool the user is using, and map to its memory/rules file:

Detection ConditionToolMemory File
~/.codebuddy/skills/ directory existsCodeBuddy.codebuddy/memory/MEMORY.md
.cursor/rules/ directory existsCursor.cursor/rules/vibe-coding-rules.mdc
.windsurfrules file existsWindsurf.windsurfrules
CLAUDE.md file existsClaude CodeCLAUDE.md
.github/copilot-instructions.md existsGitHub Copilot.github/copilot-instructions.md
None of the aboveGeneric fallback.ai-pipeline/MEMORY.md

Store detected memoryPath for use in Step -1.a and Step 6.

Step -1.a: Memory Integrity Check & Self-Heal(每次加载必执行,幂等)

  1. Read the file at memoryPath (from Step -1.0)
  2. Check for 5 critical sections (using fuzzy markdown heading match):
    • "Skill依赖链" or "Skill 编排" or "Skill Orchestration"
    • "强制自检硬规则" or "步骤6(测试)不可跳过" or "Step 6 mandatory"
    • "开发文档标准路径" or "Dev doc standard paths"
    • "面包屑规则" or "breadcrumb"
    • "项目初始化状态" or "已初始化" or "initialization status"
  3. If ANY section is missing → detect and report (V2.7: do NOT auto-write without confirmation)
  4. Report missing sections to user: "发现 Memory 文件缺少 [N] 个关键段落:[list]。是否自动补全?(是 / 跳过)"
  5. Only write to memoryPath after explicit user confirmation
  6. If file does not exist → ask: "没有检测到 Memory 文件,要创建吗?(是 / 先不要)"
  7. Record in log: self-healed: [list] or self-heal-skipped: [list]

Why V2.7 changed from auto-write to confirm / 为什么改为确认而非自动写入: Different platforms have different memory file formats (.mdc for Cursor, .codebuddy/memory/ for CodeBuddy, CLAUDE.md for Claude Code). Blind auto-writing can break platform-specific syntax or interfere with user's existing configuration. Discovery is universal; writing requires consent.

File format adaptation:

  • .mdc files (Cursor): add YAML frontmatter (---\ndescription: Vibe Coding Rules\nalwaysApply: true\n---) before markdown
  • .md / .windsurfrules / CLAUDE.md: plain markdown

This step runs on EVERY skill activation — not just first time. It ensures the pipeline's critical rules are always present in memory, preventing AI from skipping mandatory steps.

Step -1.b: Project State Detection

After memory self-heal, detect project state:

  1. Check if project root has code files (.js, .html, .py, .ts, etc.) → NO → Path A: New Project
  2. Check if 开发文档/ directory exists (or configured devDocs path) → NO → Path C: Existing Project Without Docs
  3. Check if 开发文档/00-必读/开发文档管理.md exists → YES → Path B: Existing Project With Docs → NO → Path C: Existing Project Without Docs

Path A: New Project

Proceed with standard pipeline-init workflow (Step 0 through Step 7 below).

Path B: Existing Project With Docs — Compliance Scan

  1. Run DOC_SPEC compliance scan:
    • Check all docs in 开发文档/ have 📋 管理规范见 reference in header
    • Check no orphan files in wrong directories
    • Check file size limits (≤500 lines)
    • Verify 📖 references point to existing docs
    • Verify management doc index is up to date
  2. Output a report: [N] missing 📋 references | [N] files over limit | [N] broken 📖 links
  3. Ask user: "Found some documentation issues. Fix them now or later?"
  4. If user says "fix now" → repair issues (add 📋, split large files, fix broken links)

Path C: Existing Project Without Docs — Bootstrap

V2.9 核心哲学:去模板化。 不同项目需要不同的开发文档结构,Skill 不应该拿着一套固定模板往任何项目里硬塞。正确的做法是:扫描项目 → 检测特征 → 智能提案 → 用户确认。

  1. Ask user confirmation before creating: "检测到项目缺少开发文档目录。是否自动创建? (将先扫描项目结构,根据实际技术栈和特征给出定制化建议) ① 是,帮我创建 ② 跳过(后续可手动执行「初始化开发文档」)" → If skip → exit with message: "已跳过。需要时可以说'初始化开发文档'来手动触发。" → If proceed → continue to Step 1.
  2. Language selection first: Run Phase 0 of references/init-dialog.md to ask user's preferred language (中文 / English / 日本語 / 其他).
  3. Project Scan(扫描项目特征):

    a. 文件类型分布:统计项目代码文件类型(.js/.py/.ts/.html/.css 等),了解项目主要语言。

    b. 技术栈检测(按优先级搜索):

    • 读 package.json → 检测 Web 框架(Express/Koa/Fastify)、前端框架(React/Vue/Svelte)、Electron、CLI 工具等
    • 读 requirements.txt / pyproject.toml → 检测 Python 框架(FastAPI/Django/Flask)
    • 读 Cargo.toml / go.mod / Gemfile → 检测 Rust/Go/Ruby 项目类型
    • 读 Dockerfile / docker-compose.yml / ecosystem.config.js → 部署相关
    • 读 .github/workflows/ / .gitlab-ci.yml → CI/CD 相关

    c. 关键特征检测(决定建议哪些目录):

    检测到建议目录理由
    HTTP 路由文件(routes/、server.js、app.js 等)接口文档有 API 需要文档化
    Agent/AI prompt 相关文件(prompts/、agent/、metacraft/ 等)Agent模板有 AI Agent 需要模板管理
    项目模块数 ≥ 3 且结构复杂架构设计复杂项目需架构文档
    多方案设计痕迹(proposal/、design/、方案对比文档)方案设计有方案决策需要记录
    部署配置文件(Dockerfile、k8s、PM2、Nginx 等)部署运维有部署流程需要文档
    PRD/合规/软著/法律相关文件产品与合规有合规需求
    工具/脚本集合(scripts/、tools/ 目录非空)项目参考自定义工具需要说明

    d. 不确定项标记:对于无法确定是否需要但可能相关的目录,标记为 ❓不确定,在提案中单独列出询问用户。

  4. Propose Tailored Structure(展示定制化提案):

    根据扫描结果,向用户展示提案,格式如下:

    根据项目扫描结果,建议创建以下开发文档目录:
    
    ✅ 强制(任何项目都需要):
       00-必读/          — 开发文档管理规范+项目概览
       01-变更日志/      — 每次改动的变更记录
       归档/             — 过期文档季度归档
    
    ✅ 建议创建(基于项目特征):
       04-接口文档/      — 检测到 HTTP 路由(routes/ 目录,Express 框架)
       [其他建议...]
    
    ❓ 不确定是否需要:
       06-Agent模板/     — 未检测到 AI Agent 相关代码,是否需要?
       [其他不确定...]
    
    ❌ 不建议(与项目无关):
       05-部署运维/      — 未检测到部署配置文件
       [其他不建议...]
    
    你可以:
    ① 确认全部建议(默认)
    ② 调整:添加/删除/重命名目录
    ③ 自定义:告诉我你想要的目录结构

    关键原则:

    • 永远不要让用户看到与项目无关的空目录
    • 00-必读/、01-变更日志/、归档/ 是唯一三项强制创建
    • 其他目录全部基于扫描结果建议,由用户最终决定
    • 如果用户说"我不知道需要什么"→ 展示完整可选列表,逐一询问
  5. User Confirms → Create Structure:

    用户确认后:

    • 创建用户选定的目录结构
    • 为每个目录创建 .gitkeep 或初始索引文件
    • 在 00-必读/ 中创建 开发文档管理.md,其「目录结构」节由实际创建的结构填充(不是固定模板)
  6. Auto-Populate from Scan(根据扫描结果智能填充初始内容):

    • 如果创建了 接口文档/ → 扫描路由文件,生成路由清单骨架
    • 如果创建了 Agent模板/ → 扫描 prompt 文件,列出已有模板
    • 如果创建了 架构设计/ → 生成项目文件结构概览
    • 如果创建了 项目参考/ → 列出工具脚本和配置说明
    • 所有未填充的标记为 <!-- TODO: 待补充 -->
  7. Run baseline self-check(如果项目有可执行代码):
    • 执行 self-check 周期
    • 将基线发现写入 01-变更日志/变更日志-YYYY-MM-DD.md
    • 标记为 "baseline finding — not a regression"
  8. Report: "已创建 [N] 个目录的开发文档结构。以下是初始填充内容摘要..."

Step 0: Language Selection

Follow references/init-dialog.md Phase 0 to ask the user's preferred language. All subsequent generated docs, memory files, and changelogs MUST use the selected language.

Step 1: Conversational Collection

Follow references/init-dialog.md template to progressively collect:

  1. Project name, 2. Tech stack, 3. Project type, 4. Auto-testing needed? (default yes)

Step 2: Dependency Detection

Check environment: Node.js version, npm, Git, additional deps per tech stack. On failure: provide plain-language guides. Mark missing items for later.

Step 3-5: Create Structure + Fork Template + Populate pipeline.json

Create .ai-pipeline/ with pipeline.json, VERSION (2.9.0), and self-check references. Fork self-check and changelog Skill templates. Populate project config. Write README.md with proper language and no hardcoded versions.

Step 6: Write Project Memory

Using the Step -1.0 detected platform, append V2.6 pipeline enforcement rules to the memory file, including:

  • Skill dependency chain (mandatory order)
  • Force self-check hard rules (Step 6 mandatory, Step 5.5 integrity gate)
  • Dev doc standard paths
  • Breadcrumb rules
  • Initialization status marker

Step 7: Confirm Completion

Report created structure, config items, and pipeline readiness. If memory was self-healed in Step -1.a, mention which sections were repaired.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.