三省六部技能映射

v1.0.0

将三省六部组织架构映射为可落地的多 Agent 技能协作配置。用于按中书省、门下省、尚书省与六部标签整理技能分工、职责边界、标准流程、GitHub 对象映射与数据分析分层,适合构建面向现代工程团队的三省六部协作体系。

0· 162·0 current·0 all-time
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name, description and included files all describe mapping an organizational model (三省六部) to multi‑agent skill assignments and GitHub/data objects. There are no unrelated environment variables, binaries, or install steps requested; the declared purpose matches the actual contents.
Instruction Scope
SKILL.md provides guidance for producing role mappings, workflows, and structured outputs and explicitly points to the provided YAML and publishing notes for structured data. It does not instruct reading arbitrary system files, contacting external endpoints, or accessing secrets beyond the included reference files.
Install Mechanism
No install spec and no code files — this is instruction-only. Nothing is downloaded or written to disk by the skill itself, which is the lowest-risk install posture.
Credentials
The skill requests no environment variables, credentials, or config paths. Note: while the mapping references integrations like 'github-integration', the skill itself does not include or request tokens — actual integrations would require separate skills/credentials.
Persistence & Privilege
always:false (default) and user-invocable. Autonomous invocation is permitted by platform default but the skill doesn't request elevated or persistent privileges or attempt to modify other skill/system configurations.
Assessment
This skill is essentially documentation and a template mapping and is coherent with its stated purpose. It is low-risk as provided because it has no install steps, no network endpoints, and requests no credentials. Before enabling in production: 1) Verify the skill author/owner (source is unknown) if you require provenance; 2) Remember that references to GitHub or other integrations are only mappings — actual operations against GitHub will require separate integration skills and credentials (do not put credentials into this skill); 3) If you allow autonomous agent invocation on your agent, enforce policy/guards for any downstream skills that perform network actions or need tokens; 4) Review the included YAML for any organization-specific sensitive data before sharing the skill internally.

Like a lobster shell, security has layers — review code before you run it.

latestvk974j8dw4c9rcfg2f8676wc80h83b1e0
162downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

三省六部技能映射

本 skill 用于把“三省六部 × 技能库”整理为可落地的协作框架,适用于:

  • 设计三省六部多 Agent 架构
  • 为部门分配技能与职责边界
  • 将技能映射到 GitHub、数据分析、测试、审核、文档等对象
  • 统一规划、审核、调度、执行、复盘的标准链路

使用方式

当用户要求以下任一事项时,使用本 skill:

  • 整理三省六部技能配置
  • 设计三省六部与技能库的匹配关系
  • 按中书省/门下省/尚书省/六部标签规划多 Agent 分工
  • 为古典官制风格的多 Agent 系统建立工程化协作框架
  • 输出三省六部的职责矩阵、流程图、技能映射表、配置文件

核心规则

  1. 三省负责治理,六部负责执行。
  2. 同一技能可以被多个部门使用,但必须区分主责、审核、协同。
  3. GitHub 相关能力必须映射到明确对象,如 Issue、PR、Milestone、Project、Docs、Release。
  4. 数据分析相关能力必须区分为战略分析、经营分析、执行分析。
  5. 输出结果优先沉淀为结构化配置文件,而非只给口头说明。

三省匹配

中书省

  • 定位:创意生成 → 方案撰写 → 目标制定
  • 典型技能:brainstormingwriting-plansdocxpptxpdf
  • 输出:战略摘要、方案文档、目标定义、Epic/Issue 草案

门下省

  • 定位:方案审核 → 风险校验 → 质量把关
  • 典型技能:verification-before-completionrequesting-code-reviewpdfdocx
  • 输出:审核意见、风险结论、评审报告

尚书省

  • 定位:任务分发 → 多 Agent 协同 → 进度监控
  • 典型技能:executing-plansdispatching-parallel-agentssubagent-driven-developmentmcp-builder
  • 输出:任务拆解、调度计划、执行看板、阶段汇报

六部匹配

吏部(人事 / 组织管理)

  • 典型技能:subagent-driven-developmentdispatching-parallel-agents
  • 职能:子 Agent 角色分配、权限管理、团队协作调度

户部(财务 / 数据管理)

  • 典型技能:xlsxdata-analysispdf
  • 职能:成本核算、数据统计、报表生成与可视化

礼部(品牌 / 沟通 / 文化)

  • 典型技能:brainstormingwriting-plansdocxpptxpdfcanvas-designalgorithmic-art
  • 职能:品牌宣传、活动策划、文案输出、视觉设计

兵部(项目攻坚 / 应急)

  • 典型技能:systematic-debuggingwebapp-testingexecuting-plans
  • 职能:线上问题处理、技术攻坚、应急响应

刑部(合规 / 风控 / 审核)

  • 典型技能:verification-before-completionrequesting-code-reviewxlsxpdf
  • 职能:合规检查、代码审核、风险预警、违规处理

工部(技术落地 / 开发)

  • 典型技能:test-driven-developmentsystematic-debuggingwebapp-testingfrontend-designcanvas-designalgorithmic-artweb-artifacts-buildermcp-buildersubagent-driven-developmentgithub-integration
  • 职能:前端设计、代码开发、测试部署、工程交付

标准流程

  1. 中书省:brainstormingwriting-plans → 导出 docx/pptx/pdf
  2. 门下省:verification-before-completionrequesting-code-review → 输出审核报告
  3. 尚书省:dispatching-parallel-agentsmcp-builder → 分发任务至六部
  4. 六部执行:
    • 工部:github-integrationtest-driven-developmentsystematic-debugging
    • 户部:data-analysisxlsxpdf
    • 礼部:canvas-designpptx
    • 兵部 / 刑部:全程监控质量与合规
  5. 收尾:verification-before-completion 最终校验 → 复盘 → GitHub 归档

需要时再读

如需结构化配置,读取:

  • references/sansheng_liubu_skill_map.yaml:完整 YAML 映射
  • references/publishing_notes.md:发布与维护说明

Comments

Loading comments...