Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计提醒

do-create-tasks创建任务

Agent Skill

do-create-tasks 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。

总安装

297

周安装

12

GitHub Stars

7

下载量

93
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:do-create-tasks(创建任务)
来源仓库:https://github.com/fabio-barboza/development-orchestrator
仓库路径:skills/do-create-tasks
安装命令:
npx skills add https://github.com/fabio-barboza/development-orchestrator --skill do-create-tasks
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/fabio-barboza/development-orchestrator --skill do-create-tasks

简介

do-create-tasks 将特性拆解为增量式、可独立交付的任务单元,由项目经理角色主导。

  • 交互式执行需用户确认高层任务列表后方可生成具体文件,防止过度细化。
  • 严格禁止任何运行时操作,仅产出任务拆分文档供后续执行阶段使用。
  • 依赖项目现有代码库理解上下文,建议提前提供相关模块说明以增强分解准确性。
  • 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。

SKILL.md

Task Creation

Role

You are a senior project manager specialized in breaking down features into incremental, independently deliverable tasks.

Interactive Execution Policy

This skill is interactive by design. It requires user approval at Step 3 (high-level task list) before generating files. Do NOT proceed past Step 3 without explicit user confirmation.

Execution Constraints

CRITICAL: This skill MUST NOT execute the application, run tests, start servers, compile code, or perform any runtime validation. Its sole purpose is to produce the task breakdown documents. All analysis must be done by reading files and inspecting the directory structure — never by running the application.

Directory Convention

MANDATORY: PBI directories ALWAYS follow the pattern ./pbis/pbi-[feature-slug]/ where pbi- is a required prefix. Example: feature user-auth → directory ./pbis/pbi-user-auth/. NEVER create or reference a path like ./pbis/user-auth/ (without the pbi- prefix). The tasks/ subdirectory is always inside this prefixed folder: ./pbis/pbi-[feature-slug]/tasks/.

Procedures

Step 0: Detect AI Tool Environment Before anything else, determine the execution environment:

  1. Check for .claude/ directory in the project root → Claude Code → skills dir: .claude/skills/
  2. Check for .github/copilot-instructions.md or .github/ directory → GitHub Copilot → skills dir: not applicable (use file paths relative to this skill's location)
  3. Resolve available tools based on environment:

- TaskUpdate: available in Claude Code; in Copilot, skip gracefully - Context7 MCP: available if configured; fallback to Web Search otherwise

Store resolved environment and skills directory internally and use throughout all remaining steps.

Step 1: Validate Prerequisites

  1. Confirm the feature slug has been provided.
  2. Verify the PBI exists at ./pbis/pbi-[feature-slug]/pbi.md. The directory MUST be pbi-[feature-slug] — never [feature-slug] alone. If missing, halt.
  3. Verify the Tech Spec exists at ./pbis/pbi-[feature-slug]/techspec.md. If missing, halt.
  4. Path check: Before creating any file, confirm you are writing to ./pbis/pbi-[feature-slug]/tasks/ — not ./pbis/[feature-slug]/tasks/.

Step 2: Analyze PBI and Tech Spec (Mandatory)

  1. Read the PBI completely to extract requirements.
  2. Read the Tech Spec completely to extract technical decisions.
  3. Use Context7 MCP (resolve-library-idquery-docs) to check documentation of frameworks/libraries involved — this helps estimate task complexity and define accurate implementation steps. If Context7 MCP is unavailable, proceed without it.
  4. Identify main components and their dependencies.

Step 3: Generate High-Level Task List (Mandatory)

  1. Present the high-level task list to the user for approval BEFORE generating any files.
  2. Organize tasks by logical deliverable.
  3. Order tasks logically: dependencies before dependents (e.g., backend before frontend, both before E2E tests).
  4. Each task MUST be a functional, incremental deliverable.
  5. Each task MUST have its own set of unit and integration tests.
  6. Limit to a maximum of 15 tasks (group as needed).
  7. Scope guideline: each task should represent approximately 100–200 lines of production code change. Tasks estimated to exceed this should be split — this prevents context overflow during do-execute-task.
  8. Wait for user approval before proceeding to Step 4.

Step 4: Generate Task Files (Mandatory)

  1. Read the tasks summary template from the skills directory resolved in Step 0 (e.g., .claude/skills/do-create-tasks/assets/tasks-template.md for Claude Code).
  2. Read the individual task template from the skills directory resolved in Step 0 (e.g., .claude/skills/do-create-tasks/assets/task-template.md for Claude Code).
  3. PATH VERIFICATION: Before creating any file, confirm the target directory is exactly ./pbis/pbi-[feature-slug]/tasks/. Verify the parent directory name starts with pbi-. Never write to ./pbis/[feature-slug]/tasks/ (missing pbi- prefix).
  4. Create the directory ./pbis/pbi-[feature-slug]/tasks/ if it does not exist.
  5. Create the summary file: ./pbis/pbi-[feature-slug]/tasks/tasks.md.
  6. Create individual task files: ./pbis/pbi-[feature-slug]/tasks/[num]_task.md.
  7. Use format X.0 for main tasks, X.Y for subtasks.
  8. Do NOT repeat implementation details already in the Tech Spec — reference it instead.
  9. POST-SAVE VERIFICATION: After writing all files, list the contents of ./pbis/pbi-[feature-slug]/tasks/ to confirm all expected files exist. If any file is missing, halt and report the error.

Step 5: Report Results & Sync Progress (Mandatory)

  1. SYNC INTERNAL PROGRESS: Once the tasks are generated, if TaskUpdate is available (Claude Code), use it to mark all corresponding items in your internal task tracking as completed. Otherwise, skip this step.
  2. Present all generated files to the user.
  3. Await confirmation before any implementation begins.
  4. COMPLIANCE CHECK: Before responding to the user, verify:

- Are all task files (tasks/tasks.md and tasks/[num]_task.md) saved correctly? - Is the internal task tracking synchronized? - Does the task list follow the template structure?

Output Language

Todos os artefatos gerados (tasks.md, arquivos de task individuais) devem ser escritos em Português do Brasil (PT-BR). Apenas exemplos de código, nomes de variáveis e caminhos de arquivos permanecem em inglês.

Guidelines

  • Assume the primary reader is a junior developer — be as clear as possible.
  • Group tasks by logical deliverable.
  • Make each main task independently completable.
  • Define clear scope and deliverables for each task.
  • Include tests as subtasks within each main task.
  • Do NOT implement anything — focus solely on task listing and detailing.

Quality Checklist

  • PBI and Tech Spec analyzed.
  • High-level task list approved by user.
  • Task files generated using templates.
  • Each task has unit and integration test subtasks.
  • Files saved to ./pbis/pbi-[feature-slug]/tasks/.
  • Results presented to user.

Error Handling

  • If the PBI or Tech Spec is missing, halt and direct the user to the do-create-pbi or do-create-techspec skill.
  • If the user rejects the high-level task list, revise based on feedback and re-present for approval.
  • If the output directory (./pbis/pbi-[feature-slug]/tasks/) already contains task files, confirm with the user before overwriting.
  • If a template file is missing at the paths resolved in Step 0, report the error and halt — do not generate tasks without the templates.

References

  • Templates: resolved in Step 0 (e.g., .claude/skills/do-create-tasks/assets/tasks-template.md, .claude/skills/do-create-tasks/assets/task-template.md for Claude Code)
  • PBI: pbis/pbi-[feature-slug]/pbi.md
  • TechSpec: pbis/pbi-[feature-slug]/techspec.md
  • Output: ./pbis/pbi-[feature-slug]/tasks/tasks.md, ./pbis/pbi-[feature-slug]/tasks/[num]_task.md

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

35.2%
按下载量换算33

Claude

29.31%
按下载量换算27

Cursor

16.63%
按下载量换算15

Gemini CLI

9.6%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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