Token导航 LogoToken导航

Agent Skill · 研究检索

task-breakdown-specialist(任务分解专家)

task-breakdown-specialist 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。

task-breakdown-specialist 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,权限为需要联网;许可证为 MIT,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 9,284。

总安装 14,553周安装 700GitHub Stars 26下载 7,320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
MIT

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:task-breakdown-specialist(任务分解专家)
来源仓库:https://github.com/daffy0208/ai-dev-standards
仓库路径:skills/task-breakdown-specialist
安装命令:
npx skills add https://github.com/daffy0208/ai-dev-standards --skill 'Task Breakdown Specialist'
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/daffy0208/ai-dev-standards --skill 'Task Breakdown Specialist'

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。

用于查找、检索和筛选相关信息。task-breakdown-specialist 属于研究检索类 Skill,可作为该场景下的辅助能力补充。适合根据关键词或任务场景快速定位候选结果。可结合来源仓库和原始 README 核验具体用法。安装前建议确认权限范围、维护状态及是否会触发联网或命令执行。

支持 Codex、Claude、Cursor、Gemini CLI,通过 GitHub 安装。

来源仓库内路径:skills/task-breakdown-specialist

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/daffy0208/ai-dev-standards --skill 'Task Break…;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

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

task-breakdown-specialist 怎么安装?

在终端执行 npx skills add https://github.com/daffy0208/ai-dev-standards --skill 'Task Breakdown Speci…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

task-breakdown-specialist 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。

task-breakdown-specialist 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

task-breakdown-specialist 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

task-breakdown-specialist 用的人多吗?

公开数据是累计安装 14,553 次,近一周安装 700 次,下载 7,320 次,GitHub 26 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。

task-breakdown-specialist 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 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。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 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
task-breakdown-specialist

任务分解专家 · 本页

14,55326需要联网MIT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

Task Breakdown Specialist

Lower activation energy. Build momentum. Make progress visible.

Core Principle

ADHD brains struggle with large, ambiguous tasks due to high activation energy. The solution: Break everything into tiny, achievable micro-tasks that provide frequent dopamine hits and visible progress.

Key Insight: Starting is the hardest part. Make starting ridiculously easy.


The ADHD Task Problem

Problem 1: Activation Energy

Issue: "Build authentication system" feels overwhelming Result: Procrastination, anxiety, avoidance

Problem 2: Time Blindness

Issue: Can't estimate how long tasks take Result: Unrealistic plans, missed deadlines, frustration

Problem 3: Working Memory

Issue: Can't hold large task in mind while working Result: Losing track of what you're doing, constant re-reading

Problem 4: Dopamine Deficit

Issue: Big tasks provide no intermediate rewards Result: Motivation loss, task abandonment


Phase 1: Quick Win Strategy

Start with 2-Minute Tasks

Why: Zero activation energy, instant dopamine, builds momentum

❌ Bad Task: "Build user authentication"

- Feels huge (overwhelm)
- Unclear start point (paralysis)
- No quick wins (no dopamine)

✅ Good Breakdown:

Quick Wins (Build Momentum) ⚡

1. [ ] Create /auth folder (30 sec) 🟢
2. [ ] Create auth.ts file (30 sec) 🟢
3. [ ] Install bcrypt package (1 min) 🟢

WHY START HERE:

- Combined: 2 minutes total
- Zero thinking required
- 3 quick dopamine hits
- Momentum established

THEN move to real work...

The 2-5-15-30 Pattern

Break tasks into time categories:

2-minute tasks (🟢) - Setup, scaffolding, installs 5-minute tasks (🟢) - Simple functions, basic HTML 15-minute tasks (🟡) - Core logic, API calls 30-minute tasks (🟡) - Complex features, integration

Rule: Always start with 2-minute tasks!


Phase 2: The Breakdown Formula

Step 1: Capture the Big Task

Write down the overwhelming task:

Big Task: Build user dashboard with analytics

Step 2: Ask Clarifying Questions

Questions:

- What's the absolute minimum? (MVP)
- What would "done" look like?
- What can I skip for now?
- What's P0 vs P1 vs P2?

Step 3: Identify Micro-Tasks

Use this template:

## Big Task: Build User Dashboard

### P0 (Must Have - MVP)

- Dashboard page exists
- Shows user name
- Shows one metric (login count)

### P1 (Should Have - Next)

- Charts and graphs
- Multiple metrics
- Time range selector

### P2 (Nice to Have - Later)

- Export data
- Custom widgets
- Real-time updates

Step 4: Break P0 into Micro-Tasks

## Building P0 Dashboard

Phase 1: Quick Wins (Momentum) ⚡

1. [ ] Create dashboard.tsx file (30 sec) 🟢
2. [ ] Add to navigation menu (1 min) 🟢
3. [ ] Add "Dashboard" heading (30 sec) 🟢
       Total: 2 minutes

Phase 2: Core Structure 🏗️ 4. [ ] Create dashboard layout (5 min) 🟢 5. [ ] Add user name display (5 min) 🟢 6. [ ] Style with Tailwind (10 min) 🟡
Total: 20 minutes

Phase 3: Data Integration 📊 7. [ ] Fetch user data from API (15 min) 🟡 8. [ ] Display login count (10 min) 🟡 9. [ ] Add loading state (5 min) 🟢
Total: 30 minutes

Phase 4: Polish & Ship ✨ 10. [ ] Handle error states (10 min) 🟡 11. [ ] Test all paths (10 min) 🟡 12. [ ] Deploy (5 min) 🟢
Total: 25 minutes

Grand Total: ~75 minutes
With ADHD tax (1.5x): ~110 minutes (2 hours)

Phase 3: Task Sizing with ADHD Tax

Standard Time Estimates Are Lies

Neurotypical estimate × 1.5-2 = ADHD estimate

Why:

  • Task switching overhead
  • Distraction recovery time
  • Setup/teardown time
  • "Where was I?" moments
  • Inevitable rabbit holes

ADHD Time Multipliers

Simple task (5 min estimated):
→ ADHD reality: 7-10 min (1.5x)

Medium task (30 min estimated):
→ ADHD reality: 45-60 min (1.5-2x)

Complex task (2 hours estimated):
→ ADHD reality: 3-4 hours (2x)

New/Unknown task (?? estimated):
→ ADHD reality: 3x whatever you think

Honest Task Sizing

## Feature: Add password reset

Traditional estimate: 1 hour

ADHD-honest estimate:

- Setup email service (5 min → 10 min)
- Generate reset token (10 min → 15 min)
- Build reset form (20 min → 30 min)
- Handle token validation (15 min → 25 min)
- Test flows (10 min → 20 min)

Subtotal: 60 min → 100 min

- Buffer for interruptions: +20 min
- Debugging time: +30 min

Real estimate: 2.5 hours

👍 Better to overestimate and finish early (dopamine!)
👎 Than underestimate and feel like failure

Phase 4: Momentum-Based Ordering

Dopamine-Optimized Task Sequence

Don't do tasks in logical order. Do them in motivation order.

❌ Logical Order (ADHD-hostile):

1. Plan architecture (hard, no dopamine)
2. Set up database (boring, no dopamine)
3. Build API (complex, delayed dopamine)
4. Build UI (fun but blocked by above)

✅ Momentum Order (ADHD-friendly):

1. Build UI mockup (fun, instant visual) 🎨 Dopamine!
2. Add fake data (easy, looks real) 🎯 Dopamine!
3. Hook up API (has context now) 💪 Dopamine!
4. Set up database (motivated now) ✅ Dopamine!
5. Refine architecture (understand it now) 🚀 Done!

Why this works:

- Starts with fun (low activation energy)
- Provides instant feedback (dopamine)
- Builds understanding (easier to do hard parts)
- Maintains momentum (no dead zones)

The Quick Win Sandwich

Pattern: Easy → Hard → Easy

Session Plan (2 hours):

00:00 - 00:10: Quick wins (3 easy tasks)
→ Builds momentum
→ Establishes "I'm productive" mindset

00:10 - 01:30: Hard task (focus required)
→ You have momentum now
→ Feel accomplished from quick wins

01:30 - 02:00: Easy task (cool down)
→ End on high note
→ Easy to resume tomorrow

Phase 5: Progress Visualization

Make Progress Visible

ADHD brains need to SEE progress for dopamine release.

Progress Bars
## Feature: User Authentication

Progress: ████████░░ 80%

✅ Phase 1: Setup (20%)
✅ Phase 2: Login (30%)
✅ Phase 3: Signup (30%)
🔄 Phase 4: Reset (20%) ← YOU ARE HERE

Next: Password reset form (15 min)
Checklist with Time Tracking
## Today's Tasks

🔥 Streak: 3 days
⏱️ Focus time: 45 min

Morning Session (2 hours budgeted):

- [x] Fix login bug (15 min) ✅ 12 min (faster!)
- [x] Add loading spinner (5 min) ✅ 7 min
- [x] Update tests (10 min) ✅ 15 min
- [ ] Code review (20 min) ← CURRENT
- [ ] Deploy fix (10 min)

Progress: 3/5 tasks (60%)
Time spent: 34 min / 120 min
Remaining: 30 min of work

💪 You're crushing it!
Commit-Based Progress
## This Week's Progress

Mon: ✅✅✅✅✅ (5 commits)
Tue: ✅✅✅ (3 commits)
Wed: ✅✅✅✅✅✅✅ (7 commits) 🔥 Best day!
Thu: ✅✅✅✅ (4 commits)
Fri: ✅ (1 commit) ← TODAY

Total: 20 commits this week
Goal: 15 commits (exceeded! 🎉)

Keep the streak alive!

Phase 6: Task Breakdown Patterns

Pattern 1: Feature Breakdown

# Feature: Add dark mode

## Step 1: Identify Components

What needs to change?

- Colors
- Component styles
- Toggle button
- Persistence

## Step 2: Break Down Each

### Colors (30 min total)

- [ ] Define dark color palette (5 min) 🟢
- [ ] Add CSS variables (10 min) 🟡
- [ ] Test contrast ratios (15 min) 🟡

### Component Styles (45 min total)

- [ ] Update Button component (10 min) 🟡
- [ ] Update Card component (10 min) 🟡
- [ ] Update Input component (10 min) 🟡
- [ ] Update Header component (15 min) 🟡

### Toggle Button (20 min total)

- [ ] Create ThemeToggle component (15 min) 🟡
- [ ] Add to header (5 min) 🟢

### Persistence (15 min total)

- [ ] Save preference to localStorage (10 min) 🟡
- [ ] Load on app start (5 min) 🟢

Total: 110 min (2 hours realistic)

Pattern 2: Bug Fix Breakdown

# Bug: Users can't reset password

## Step 1: Reproduce (10 min)

- [ ] Try reset flow locally (5 min) 🟢
- [ ] Check error logs (5 min) 🟢

## Step 2: Investigate (20 min)

- [ ] Find reset-password.ts file (1 min) 🟢
- [ ] Read through code (5 min) 🟢
- [ ] Identify issue (token expiry) (10 min) 🟡
- [ ] Confirm root cause (4 min) 🟢

## Step 3: Fix (30 min)

- [ ] Update token expiry logic (20 min) 🟡
- [ ] Add better error message (10 min) 🟡

## Step 4: Test (15 min)

- [ ] Test locally (10 min) 🟡
- [ ] Test on staging (5 min) 🟢

## Step 5: Deploy (10 min)

- [ ] Create PR (3 min) 🟢
- [ ] Deploy to production (7 min) 🟢

Total: 85 min (1.5 hours)

Pattern 3: Learning Task Breakdown

# Learn: Next.js App Router

❌ Bad: "Learn Next.js App Router" (too vague)

✅ Good Breakdown:

## Phase 1: Overview (30 min)

- [ ] Watch official intro video (15 min) 🟢
- [ ] Read "Getting Started" docs (15 min) 🟢

## Phase 2: Hands-On (1 hour)

- [ ] Create new Next.js project (5 min) 🟢
- [ ] Build simple page (15 min) 🟡
- [ ] Add dynamic route (20 min) 🟡
- [ ] Try server components (20 min) 🟡

## Phase 3: Apply (2 hours)

- [ ] Migrate one page in my project (30 min) 🟡
- [ ] Test and debug (30 min) 🟡
- [ ] Migrate second page (30 min) 🟡
- [ ] Document learnings (30 min) 🟡

Total: 3.5 hours over 2 days
Day 1: Phase 1 + 2 (1.5 hours)
Day 2: Phase 3 (2 hours)

Phase 7: Anti-Patterns (What NOT to Do)

❌ Anti-Pattern 1: Perfectionist Breakdown

Bad:

- [ ] Research best authentication library (no time limit)
- [ ] Compare 10 different approaches (rabbit hole)
- [ ] Design perfect architecture (never starts)

Good:

- [ ] Choose auth library (10 min, use NextAuth)
- [ ] Follow quickstart guide (30 min)
- [ ] Ship MVP, iterate later

❌ Anti-Pattern 2: Too Granular

Bad:

- [ ] Open VS Code (10 sec)
- [ ] Navigate to file (10 sec)
- [ ] Type 'import' (5 sec)

This is ridiculous. You'll spend more time reading the list.

Good:

- [ ] Add imports to auth.ts (2 min)

❌ Anti-Pattern 3: All Hard Tasks

Bad:

- [ ] Refactor authentication (2 hours)
- [ ] Optimize database queries (3 hours)
- [ ] Fix critical bug (2 hours)

No quick wins = No momentum = No dopamine = Procrastination

Good:

- [ ] Update README (5 min) 🟢 ← Start here!
- [ ] Refactor one function (30 min) 🟡
- [ ] Take break (5 min)
- [ ] Fix critical bug (2 hours) 🟡

❌ Anti-Pattern 4: Vague Tasks

Bad:

- [ ] Work on dashboard
- [ ] Fix bugs
- [ ] Improve performance

What does "done" mean? ADHD brain will wander.

Good:

- [ ] Add user stats to dashboard (15 min)
- [ ] Fix login timeout bug (#123) (20 min)
- [ ] Reduce API response time to < 200ms (1 hour)

Phase 8: Emergency Task Breakdown

When You're Stuck RIGHT NOW

5-Minute Breakdown Exercise:

  1. Write the overwhelming task: _________________________________
  2. What's the FIRST tiny action? (< 2 min) _________________________________
  3. Do ONLY that action. Nothing else.
  4. Dopamine hit! Now write next action: _________________________________
  5. Repeat.

Example:

1. Overwhelming: "Build entire checkout flow"
2. First tiny action: "Create checkout.tsx file"
3. [DO IT NOW - 30 seconds]
4. ✅ Done! Dopamine!
5. Next action: "Add basic form HTML"
6. [DO IT - 5 min]
7. Keep going...

Checklist: Good Task Breakdown

Use this to validate your task breakdown:

  • Tasks are 15 minutes or less
  • Starts with 2-3 "quick win" tasks (2-5 min)
  • Each task has clear "done" criteria
  • Time estimates include ADHD tax (1.5x)
  • Uses progress visualization (%, bars, checklists)
  • Easy tasks mixed with hard tasks (momentum maintenance)
  • No vague tasks ("work on", "improve", "fix")
  • Total time estimated honestly (not wishful thinking)
  • First task has near-zero activation energy

Tools & Templates

Daily Task Template

# Today: [Date]

## ONE Main Goal

[The ONE thing that matters today]

## Morning Session (2 hours)

Quick Wins:

- [ ] Task 1 (2 min) 🟢
- [ ] Task 2 (5 min) 🟢

Main Work:

- [ ] Task 3 (30 min) 🟡
- [ ] Task 4 (45 min) 🟡

Cool Down:

- [ ] Task 5 (10 min) 🟢

## Afternoon Session (2 hours)

[Same pattern]

## Wins Log

[What you completed - for dopamine]

Project Breakdown Template

# Project: [Name]

## MVP Definition

What's the absolute minimum to ship?

## Phases

### Phase 1: Foundation (Quick wins)

- [ ] Tasks...

### Phase 2: Core Features

- [ ] Tasks...

### Phase 3: Polish

- [ ] Tasks...

## Time Estimate

Naive: X hours
ADHD-realistic: Y hours (1.5-2x)

Related Skills

  • context-preserver - Save state between tasks
  • focus-session-manager - Manage work sessions
  • completion-coach - Finish what you start
  • adhd-workflow-architect - Design ADHD-friendly workflows

Success Metrics

You're using this skill well when:

  • ✅ Starting tasks feels easy (low activation energy)
  • ✅ You see progress every 15 minutes
  • ✅ Time estimates are realistic
  • ✅ You finish tasks instead of abandoning them
  • ✅ You feel in control (not overwhelmed)

Break it down. Start small. Build momentum. Ship things. 💪

平台分布

Claude Code

26.21%
按下载量换算1,919

OpenCode

22.66%
按下载量换算1,659

Antigravity

17.75%
按下载量换算1,299

Gemini CLI

12.06%
按下载量换算883

Codex

7.24%
按下载量换算530

Cursor

3.7%
按下载量换算271

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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。

继续浏览同类 Skills