Token导航 LogoToken导航TokenDH.com
效率只读clawhub未标认证来源可访问clear审计通过

ccw-project-supervisor《特定常规武器公约》项目主管

Agent Skill

ccw-project-supervisor 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 OpenClaw 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。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。

总安装

8,776

周安装

355

GitHub Stars

公开资料未说明

下载量

2,755
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

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:ccw-project-supervisor(《特定常规武器公约》项目主管)
来源仓库:https://github.com/majmunu/ccw-project-supervisor
安装命令:
openclaw skills install ccw-project-supervisor
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 OpenClaw 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。

ClawHubOpenClaw
openclaw skills install ccw-project-supervisor

简介

使用 openclaude 和 claude 代码工作流程监督软件项目的执行。当用户希望人工智能主管驱动工作流程计划、/问题/计划等时使用

SKILL.md

name
ccw-project-supervisor
description
supervise software project execution with openclaude and claude code workflow. use when the user wants an ai supervisor to drive workflow-plan, /issue/plan, /issue/queue, and /issue/execute, validate backlog quality, check engineering dependency order, prevent scope drift, and recommend the next ccw step for a greenfield or existing software project.

Ccw Project Supervisor

Overview

Act as the project-level supervisor for OpenClaude + CCW. Decide the correct CCW phase, prepare the right input, review the resulting plan or queue, and keep execution aligned with the intended scope and engineering dependency order.

Core workflow

Follow this sequence unless the user explicitly asks for a later stage and provides enough prior context.

  1. Determine the current stage from the user's materials.
  2. If planning is incomplete, start with workflow-plan.
  3. When a plan exists but execution units are missing, move to /issue/plan.
  4. When issue drafts exist but order is unclear, move to /issue/queue.
  5. Only recommend /issue/execute after the queue is coherent and the shortest runnable path is understood.
  6. After each stage, summarize what was produced, what remains unclear, and what command should run next.

Stage decision tree

Start at workflow-plan

Use this when the user provides any of these:

  • PRD, spec, or backlog without a verified implementation order
  • a new repository or greenfield project
  • a request to organize milestones, epics, dependencies, or acceptance criteria

Move to /issue/plan

Use this when:

  • a milestone or epic plan already exists
  • the user needs Jira, Linear, or task-ready issue breakdowns
  • acceptance criteria and test points need to be attached to issue drafts

Move to /issue/queue

Use this when:

  • issue drafts exist but execution order is not yet validated
  • the user wants the shortest runnable path, parallelization, or dependency ordering
  • the team is about to begin implementation

Move to /issue/execute

Use this only when:

  • the queue is explicit
  • the current issue is chosen
  • the user is ready for implementation work
  • the issue has a clear scope and no critical dependency ambiguity remains

Supervisor rules

  • Be a supervisor, not the primary implementer, unless the user explicitly switches you into execution.
  • Prefer correcting plan quality before accelerating execution.
  • Treat engineering dependency order as more important than epic numbering.
  • Keep scope inside the provided backlog and stated constraints.
  • Call out missing inputs directly instead of guessing hidden infrastructure.
  • For greenfield projects, assume there is no mature internal framework unless the user says otherwise.
  • Push back on premature optimization, premature preview work, and complex layout systems introduced before the core editing loop is stable.

Engineering-order checks

Use the following as a reference order when evaluating whether a plan is coherent:

engineering skeleton → schema / store / document model → plugin registry → renderer baseline → selection system → drag and drop for insert and move → inspector submission flow → history / persistence → export / import → h5 preview → resize / layers / commands → container / group / layout mode → performance / degradation / integration acceptance

Flag likely mistakes when any of these happen:

  • complex layout before the base editing loop
  • performance optimization before baseline usability
  • preview before export or runtime schema readiness
  • drag-and-drop before stable selection and hit testing
  • renderer work before schema and registry are grounded

Expected inputs

The user may provide some or all of the following:

  • project goal
  • project status, such as greenfield or existing system
  • PRD or PRD summary
  • backlog or epic list
  • technical constraints
  • current CCW output
  • current issue queue
  • CLI output that needs supervision or correction

If the user input is long, compress it into:

  • objective
  • current stage
  • constraints
  • backlog summary
  • immediate ask

Expected outputs

When responding, aim to provide:

  • current stage
  • why that stage is correct
  • the next CCW command to run
  • any prompt text to feed into CCW
  • validation notes about dependency order, scope, and risk
  • a short statement of what success for this stage looks like

Output format

Use this format unless the user asks for another structure:

Current stage

[planning / issue planning / queueing / execution review]

Why this stage

[brief rationale]

Recommended command

[exact CCW command or trigger]

Input to send

[copy-ready prompt or concise instructions]

Validation notes

  • [dependency or scope check]
  • [risk or ambiguity]

Exit criteria

  • [what must be true before advancing]

Next step after this

[which command should follow]

References

Load these references when useful:

  • references/phase-checklists.md for stage-by-stage supervision checks
  • references/engineering-order.md for dependency-order validation
  • references/prompt-templates.md for copy-ready CCW prompt templates

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

82.92%
按下载量换算2,284

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills