Token导航 LogoToken导航TokenDH.com
待分类只读github未标认证来源可访问许可证需确认审计提醒

planning规划

Agent Skill

planning 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 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

470

周安装

20

GitHub Stars

公开资料未说明

下载量

165
CodexClaudeCursorGemini CLI

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

unknown

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:planning(规划)
来源仓库:https://github.com/caldiaworks/caldiaworks-marketplace
仓库路径:skills/planning
安装命令:
npx skills add https://github.com/caldiaworks/caldiaworks-marketplace --skill planning
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/caldiaworks/caldiaworks-marketplace --skill planning

简介

planning 分析 GitHub Issues 依赖关系并输出持久化执行计划文档。

  • 适用于多 Issue 协同开发与任务排序场景。
  • 通过 npx skills add 命令从 caldiaworks 市场安装,递归读取父子 Issue 结构。
  • 输出包含层级树与标签摘要,帮助用户理解整体工作流。
  • planning 属于待分类类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Work Planning

Analyze a set of GitHub Issues, identify dependencies, propose an execution order, and define work units. Output a persistent plan document that survives context loss.

Works in any project with GitHub Issues.

Workflow

Step 1: Analyze Issues

  1. Accept a parent issue number from the user (or a list of issue numbers).
  2. Read the issue hierarchy using gh sub-issue list <number> recursively.
  3. For each issue, read the title, labels, body, and parent-child relationships.
  4. Display a summary of all issues to the user:

- Issue number and title - Labels - Parent-child hierarchy (indented tree)

If reading issues from GitHub fails, display the error message and stop. If the specified parent issue has no sub-issues, inform the user there are no issues to plan.

Step 2: Identify Dependencies and Determine Order

  1. Analyze dependencies between issues based on:

- Explicit references in issue bodies (e.g., "depends on #X", "blocked by #X") - Hierarchy: parent issues depend on child issues being resolved - Content: issues that produce artifacts needed by other issues

  1. Determine an execution order based on dependencies and priorities.

When determining order, consider practical benefits: foundational tools first (e.g., a commit skill before other skills, so it can be used during their development).

Step 3: Define Work Units

Group related issues into work units. Each work unit:

  • Corresponds to a single feature branch
  • Contains issues that are closely related and make sense to implement together
  • Has a clear, concise description

A work unit may contain a single issue or multiple related issues.

Step 4: Output Plan Document

Generate a Markdown plan document and save it to .docs/plans/.

The document contains:

# Plan: {Title}

- Date: {YYYY-MM-DD}
- Parent Issue: #{number}
- Requirements Document: {REQ-DOC-ID, if applicable}

## Dependencies

| REQ/Issue | Description | Depends On |
|-----------|-------------|------------|
| #{number} | {title} | {dependencies} |

## Execution Order

| Order | Work Unit | Issue(s) | Branch | Status |
|-------|-----------|----------|--------|--------|
| 1 | {description} | #{number} | TBD | Pending |

## Rationale

- **Order 1 ({description})**: {why this order}

## Next Action

Work Unit 1: {description} (#{number}).

After the plan is finalized, display the first work unit and its associated issues as the next action.

Before Finishing

After generating the plan document, verify it stands on its own:

  • Survives context loss? If the conversation is lost and only the plan file remains, can someone resume work from it? Every work unit should reference concrete issue numbers, not just descriptions.
  • Actionable? Does each work unit have enough detail to create a branch and start working without re-reading the full issue thread?

Constraints

  • Do not modify issues (no labels, assignees, or status changes) -- planning is read-only.
  • Do not create branches or start implementation -- planning produces a document only.
  • Always persist the plan to a file so it survives context loss.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

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

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

Codex

34.26%
按下载量换算57

Claude

29.41%
按下载量换算49

Cursor

18.97%
按下载量换算31

Gemini CLI

10.53%
按下载量换算17

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。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