Token导航 LogoToken导航

Agent Skill · 研究检索

main-workflow-router(主工作流程路由器)

main-workflow-router 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

main-workflow-router 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为权限未标注;许可证信息还不明确,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 58,594。

总安装 412周安装 17GitHub Stars 67下载 135核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:main-workflow-router(主工作流程路由器)
来源仓库:https://github.com/bacoco/bmad-skills
仓库路径:skills/main-workflow-router
安装命令:
npx skills add https://github.com/bacoco/bmad-skills --skill main-workflow-router
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/bacoco/bmad-skills --skill main-workflow-router

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

main-workflow-router 用于查找和筛选相关信息。适合在关键词搜索或任务场景中快速定位候选结果。可结合来源仓库和原始 README 核验具体用法。安装前应确认权限范围、维护状态及是否触发联网或文件读写。

适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/main-workflow-router

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/bacoco/bmad-skills --skill main-workflow-route…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    权限没有明确标注,第一次使用时只给一个无关紧要的任务,观察它到底读了哪些文件和网络。

常见问题

6 条

main-workflow-router 怎么安装?

在终端执行 npx skills add https://github.com/bacoco/bmad-skills --skill main-workflow-router。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

main-workflow-router 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

main-workflow-router 需要什么权限?

权限级别没有明确标注。第一次使用时建议只给无关紧要的任务,观察它实际读写了哪些内容。

main-workflow-router 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

main-workflow-router 用的人多吗?

公开数据是累计安装 412 次,近一周安装 17 次,下载 135 次,GitHub 67 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

main-workflow-router 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
main-workflow-router

主工作流程路由器 · 本页

41267权限未标注未明确未通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

End-to-End Orchestration Skill

When to Invoke

ALWAYS auto-invoke at the start of any BMAD project:

  • User says "start project", "new project", "initialize BMAD", "begin"
  • User says "what's next?", "where am I?", "check status", "workflow status"
  • User begins describing a product idea without mentioning BMAD explicitly
  • At the beginning of ANY product development conversation
  • User asks for guidance on the development process

Special auto-behaviors:

  • If no workflow-status.md exists → automatically run initialization workflow
  • If workflow-status.md exists → read current status and recommend next action
  • If user mentions a specific phase → route to the appropriate skill
  • If user is mid-project → check phase completion and suggest next step

Routing intelligence based on user intent:

FIRST: Assess complexity level (0-4)

  • Level 0-1 (bug fix, small change, simple feature) → Route to OpenSpec
  • Level 2-4 (new product, complex feature, architecture) → Route to BMAD

For OpenSpec work (Level 0-1):

  • Mentions "bug", "fix", "small change", "quick" → openspec-change-proposal
  • Has proposal, says "implement", "apply" → openspec-change-implementation
  • Change complete, says "archive", "close" → openspec-change-closure

For BMAD work (Level 2-4):

  • Mentions "idea", "brainstorm", "explore" → bmad-discovery-research
  • Mentions "PRD", "requirements", "plan" → bmad-product-planning
  • Mentions "UX", "UI", "design" → bmad-ux-design
  • Mentions "architecture", "tech stack", "build" → bmad-architecture-design
  • Mentions "test strategy", "QA", "ATDD" → bmad-test-strategy
  • Mentions "stories", "breakdown", "backlog" → bmad-story-planning
  • Mentions "implement story", "code", "develop" → bmad-development-execution

Critical decision: OpenSpec vs BMAD

  • Small, well-defined, existing codebase → OpenSpec
  • Greenfield, complex, or needs discovery → BMAD
  • When in doubt, ask user about complexity level

Do NOT invoke when:

  • User is clearly asking for a specific skill (let that skill handle it)
  • User is in the middle of implementing code (appropriate skill is active)
  • User is asking technical questions unrelated to workflow

Auto-Activation Behavior

This skill applies specialized routing heuristics whenever it auto-invokes:

1. Conversation Start Detection

When a conversation begins and the user mentions:

  • Product development topics
  • Project ideas or features
  • Building or creating something
  • BMAD or OpenSpec methodology

Auto-action: Initialize workflow files if none exist and recommend the right path.

2. Status Check Detection

When user asks:

  • "Where am I?"
  • "What's next?"
  • "What should I do now?"
  • "Check workflow status"

Auto-action: Read workflow-status.md, summarize phase progress, and suggest next skill.

3. Phase Transition Detection

When a downstream skill completes its phase:

  • Analyst delivers discovery brief → confirm Planning readiness
  • PM delivers PRD/epics → prepare Solutioning
  • Architecture delivers decision doc → move to Implementation

Auto-action: Update workflow status and queue the appropriate next capability.

4. Routing Intelligence

Based on user intent, automatically route to:

  • Mentions "idea", "brainstorm" → bmad-discovery-research
  • Mentions "PRD", "requirements" → bmad-product-planning
  • Mentions "UX", "UI" → bmad-ux-design
  • Mentions "architecture", "tech stack" → bmad-architecture-design
  • Mentions "test strategy", "QA" → bmad-test-strategy
  • Mentions "stories", "backlog" → bmad-story-planning
  • Mentions "implement", "code" → bmad-development-execution
  • Mentions "small change", "bug" → openspec-change-proposal

5. Safety Checks

Before routing, verify:

  • Required artifacts from previous phase exist
  • Quality gates in CHECKLIST.md are satisfied
  • No unresolved blockers in workflow-status.md

If checks fail, halt routing and request the missing prerequisites.

Mission

Serve as the global orchestrator for all product development work, intelligently routing between BMAD (Level 2-4 complex projects) and OpenSpec (Level 0-1 lightweight changes). Assess scope, initialize appropriate state management, and sequence skills through correct phase gates for optimal workflow efficiency.

Inputs Required

  • project_summary: current objective, level, and stakeholder context
  • artifacts_index: list of delivered files and their status
  • status_files: docs/bmad-workflow-status.md, docs/sprint-status.yaml, or equivalents if they exist

Outputs

  • workflow-status.md - Project status tracking document (generated from assets/workflow-status-template.md.template)
  • Updated workflow and sprint status records via scripts/workflow_status.py and scripts/sprint_status.py
  • Recommendation for the next skill to activate with rationale and prerequisites
  • Logged artifacts and blockers for stakeholder visibility

Template location: .claude/skills/main-workflow-router/assets/workflow-status-template.md.template

Process

  1. Assess Complexity & Choose Workflow

- Determine project level (0-4) from user description - Level 0-1: Route to OpenSpec (propose → implement → archive) - Level 2-4: Route to BMAD (analysis → planning → solutioning → implementation) - When ambiguous: Ask clarifying questions about scope and complexity

  1. Initialize State Management

- For BMAD: Create/update docs/bmad-workflow-status.md and docs/sprint-status.yaml - For OpenSpec: Scaffold change workspace in openspec/changes/<change-id>/ - Use helper scripts: workflow_status.py, sprint_status.py, scaffold_change.py

  1. Review Context & Prerequisites

- Check existing artifacts from previous phases - Validate gate criteria from preceding skills - Identify missing prerequisites or blockers

  1. Route to Next Skill

- For OpenSpec: openspec-change-proposal → openspec-change-implementation → openspec-change-closure - For BMAD: Route through phases (analyst → pm → ux/architecture/tea → stories → dev) - Communicate required inputs and outstanding risks

  1. Update State & Report Progress

- Update workflow status files - Log completed artifacts and next actions - Highlight blockers and recommend remediation - Escalate from OpenSpec to BMAD if scope grows beyond Level 1

Quality Gates

All items in CHECKLIST.md must be satisfied to progress between phases. Never advance without required artifacts.

Error Handling

  • When status files are absent, run initialization scripts and request missing information.
  • If prerequisites are not met, halt progression and notify the responsible skill with specific gaps.
  • Downgrade to lightweight workflows (e.g., OpenSpec) when project level is 0-1 and BMAD overhead is unnecessary.

平台分布

Codex

37.19%
按下载量换算50

Claude

26.02%
按下载量换算35

Cursor

19.12%
按下载量换算26

Gemini CLI

9.63%
按下载量换算13

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

通过

权限和风险

权限未标注

当前来源未能明确判断权限范围,默认进入异常复核队列。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills